ARTICLE

THE MYTH OF THE NONDELEGATION DOCTRINE

KEITH E. WHITTINGTON† & JASON IULIANO††

For much of the nineteenth and early twentieth centuries, the nondelegation doctrine served as a robust check on governmental expansion. Then, during the New Deal revolution, the Supreme Court reined in the doctrine, thereby paving the way for the rise of the modern administrative state. This story is one we all know well. It is taught in every constitutional law class and has been endorsed by constitutional law scholars since the 1930s. In this Article, we are the first to challenge this narrative. Our investigation draws upon an original dataset we compiled that includes every federal and state nondelegation challenge before 1940—more than two thousand cases in total. In reviewing these judicial decisions, we find that the nondelegation doctrine never actually constrained expansive delegations of power. Ultimately, our analysis reveals that the traditional narrative behind the nondelegation doctrine is nothing more than a myth.

INTRODUCTION ...... 380 I. THE “CONSTITUTION IN EXILE” ...... 384 II. DEVELOPMENT OF THE NONDELEGATION DOCTRINE...... 388 A. The Constitutional Foundations ...... 388 B. The Doctrine in the Supreme Court ...... 392 C. The Doctrine in State and Lower Federal Courts ...... 405 III. JUDICIAL PRACTICE IN NONDELEGATION CASES ...... 417 CONCLUSION ...... 429

† William Nelson Cromwell Professor of Politics, Princeton University. †† Associate Research Scholar, Yale Law School; Ph.D. candidate in Politics, Princeton University; J.D., Harvard Law School. We thank Richard Epstein, Samuel Estreicher, Roger Ford, Jacob Gersen, and Daniel Kelly for valuable comments and discussions relating to this project.

(379) 380 University of Pennsylvania Law Review [Vol. 165: 379

INTRODUCTION Nondelegation doctrine cases follow a predictable pattern. Every few years, a court of appeals invokes the doctrine to strike down a federal statute.1 The Supreme Court inevitably grants certiorari and overturns the appellate decision, holding that the statute is a constitutional delegation of legislative authority.2 Without fail, the Supreme Court ruling sparks a heated debate among constitutional law scholars. Some support the decision—praising the Court for driving another nail into the coffin of the nondelegation doctrine.3 Others oppose the decision—lamenting the Court’s failure to revive the doctrine and use it to rein in the ever-expanding administrative state.4 Despite their intense disagreement over what role the nondelegation doctrine ought to play in today’s legal system, both groups of scholars agree on the role that the doctrine has played throughout U.S. history. Specifically, they all endorse the narrative that, during the nineteenth and early twentienth centuries, the nondelegation doctrine served as a meaningful check on the unbridled expansion of the administrative state.5 Then, during the New Deal, the Supreme Court dismantled the doctrine and paved the way for Congress to delegate away any powers it deemed appropriate.6

1 See, e.g., Ass’n of Am. R.Rs. v. U.S. Dep’t of Transp., 721 F.3d 666, 677 (D.C. Cir. 2013) (holding that the Passenger Railroad Investment and Improvement Act of 2008 unconstitutionally “delegate[d] regulatory authority” to Amtrak); Am. Trucking Ass’ns v. EPA, 175 F.3d 1027, 1038 (D.C. Cir. 1999) (finding that the EPA relied on a construction of the Clean Air Act in promulgating [air quality standards] that “involve[d] an unconstitutional delegation of [legislative] power”). 2 See, e.g., Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1233-34 (2015) (overturning the D.C. Circuit’s decision regarding the unconstitutionality of the Passenger Railroad Investment and Improvement Act of 2008’s delegation of authority); Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 474 (2001) (finding that “[t]he scope of discretion [the Clean Air Act provision in question] allows is in fact well within the outer limits of our nondelegation precedents”). 3 See Cass R. Sunstein, Constitutionalism After the New Deal, 101 HARV. L. REV. 421, 494 (1987) (arguing that “[a] general revival of the nondelegation doctrine would also be a mistake in light of a range of considerations,” including the reasonableness of delegation and the difficulty judges would have in managing a nondelegation principle). 4 See Peter H. Aranson et al., A Theory of Legislative Delegation, 68 CORNELL L. REV. 1, 8-17, 63-67 (1982) (tracing the roots of the delegation principle and arguing for a “renewed nondelegation doctrine”); Gary Lawson, Delegation and Original Meaning, 88 VA. L. REV. 327, 335-43 (2002) (defending the nondelegation doctrine on originalist grounds); David Schoenbrod, The Delegation Doctrine: Could the Court Give It Substance?, 83 MICH. L. REV. 1223, 1249-74 (1985) (proposing an improved test for determining what congressional delegations of power are improper). 5 See Sunstein, supra note 3, at 447 (noting the opposition of the courts to “statutes creating agencies” on nondelegation grounds during the early New Deal era); see also Aranson et al., supra note 4, at 8-17 (describing the traditional nondelegation narrative of the Supreme Court using the doctrine to prevent the expansion of federal New Deal programs). 6 See Gary Lawson, The Rise and Rise of the Administrative State, 107 HARV. L. REV. 1231, 1237-41 (1994) (arguing that the Supreme Court has abandoned the nondelegation doctrine as a tool to hold legislation unconstitutional); Sunstein, supra note 3, at 447-48, 482 (noting “the downfall of the nondelegation doctrine” when the Court ended its assault on New Deal reforms and began tolerating broad delegations of authority). 2017] The Myth of the Nondelegation Doctrine 381

In this Article, we argue that this narrative is wrong. Drawing from our own dataset of more than two thousand nondelegation cases, we show that there was never a time in which the courts used the nondelegation doctrine to limit legislative delegations of power. In short, we expose the myth of the nondelegation doctrine. Before presenting these findings, we briefly situate the nondelegation doctrine in the context of the New Deal to illustrate why our analysis represents such an important and dramatic departure from accepted wisdom. As scholars have long maintained, the New Deal transformed the landscape of American constitutional law. The constitutional scholar Edward Corwin characterized the U.S. Supreme Court’s capitulation to the Franklin D. Roosevelt Administration as a “constitutional revolution.”7 Historian William Leuchtenberg saw the Supreme Court—and the Constitution—as “reborn” following the 1937 “switch in time.”8 Constitutional theorist Bruce Ackerman characterized the New Deal as a moment of transformation that marked the transition to an entirely new constitutional regime.9 And political scientist Howard Gillman concluded that the Constitution, as well-trained lawyers understood it, was “besieged” and eventually “collapsed” in the 1930s.10 Part of what distinguishes the New Deal from other episodes of constitutional change is the breadth and depth of the reevaluation of established constitutional doctrine. Most famously, the bonds of “substantive due process” and expansive constitutional protections for property rights were cast off in favor of a regime of “preferred freedoms.”11 The economic

7 See EDWARD S. CORWIN, CONSTITUTIONAL REVOLUTION, LTD. 112-14 (Am. Offset Printers rev. ed. 1946) (1941) [hereinafter CORWIN, CONSTITUTIONAL REVOLUTION] (arguing that the Court’s decisions upholding New Deal legislation greatly weakened the power of judicial review to check “national legislative power”). 8 See WILLIAM E. LEUCHTENBURG, THE SUPREME COURT REBORN: THE CONSTITUTIONAL REVOLUTION IN THE AGE OF ROOSEVELT 233, 235 (1995) (noting the consensus among historians that the year 1937 marked the beginning of a new era in the Supreme Court’s constitutional jurisprudence and “altered fundamentally the character of the Court’s business, the nature of its decisions, and the alignment of its friends and foes”). 9 See 1 BRUCE ACKERMAN, WE THE PEOPLE: FOUNDATIONS 114-15 (1991) (describing the Court’s shift during the New Deal as “a transformative [constitutional] amendment expressing a profound, but not total, change in American constitutional identity”). 10 See HOWARD GILLMAN, THE CONSTITUTION BESIEGED: THE RISE AND DEMISE OF LOCHNER ERA POLICE POWERS JURISPRUDENCE 201 (1993) (“The ‘constitutional revolution of 1937’ marked the moment when the Founders’ conception of a faction-free American Republic collapsed . . . .”). 11 See G. EDWARD WHITE, THE CONSTITUTION AND THE NEW DEAL 198-99 (2000) (describing the traditional account of the Supreme Court’s retreat in political economy cases involving the Commerce, Due Process, and Contract Clauses); Howard Gillman, Preferred Freedoms: The Progressive Expansion of State Power and the Rise of Modern Civil Liberties Jurisprudence, 47 POL. RES. Q. 623, 625 (1994) (arguing that as the courts relaxed restrictions on the legislative power to regulate the economy, they enshrined individual liberties as “preferred freedoms” that would receive special judicial protection). 382 University of Pennsylvania Law Review [Vol. 165: 379 rights jurisprudence of Lochner12 gave way to the individual and political rights jurisprudence of Carolene Products.13 At the same time, the constitutional rules of federalism were thoroughly reevaluated to empower the national government to take more action to guide social and economic development within the nation.14 The scope of federal regulatory power was dramatically expanded,15 while federal budgetary authority was given new significance.16 Another key component of the New Deal constitutional revolution was the alteration of the separation of powers and the expansion of presidential power. In particular, the Court significantly reworked the nondelegation doctrine in order to allow the assignment of substantial discretion over regulatory policy to executive branch officials.17 Reviled by some, the Court’s new approach to the delegation of legislative power underscored its reassessment of the judicial role within the constitutional system that was visible in other areas of constitutional law.18 Henceforth, the federal judiciary was to take a hands-off approach to assessing thecongressional assignment of policy responsibility to other government officials. In the view of some critics, the Court transformed the constitutional scheme into a system of executive governance. As Judge Douglas Ginsburg famously put the issue, [F]or 60 years the nondelegation doctrine has existed only as part of the Constitution-in-exile . . . . The memory of these ancient exiles, banished for standing in opposition to unlimited government, is kept alive by a few

12 Lochner v. New York, 198 U.S. 45 (1905). 13 United States v. Carolene Prods. Co., 304 U.S. 144 (1938). 14 See ACKERMAN, supra note 9, at 121 (“For the New Deal Justices, it was clear that the 1930’s had swept away the old law of economic relations . . . .”). 15 See, e.g., Wickard v. Filburn, 317 U.S. 111, 124-25 (1942) (interpreting the Commerce Clause to permit the federal government’s regulation of wheat grown for home consumption because the wheat could have a “substantial economic effect oninterstate commerce”); United States v. Darby, 312 U.S. 100, 114-15 (1941) (holding that the Commerce Clause grants Congress the power to regulate employment conditions); NLRB v. Jones & Laughlin Steel Corp., 301 U.S. 1, 36-38 (1937) (finding that the Commerce Clause grants Congress the power to regulate intrastate labor relations if they have a “close and substantial relation to interstate commerce”). 16 See, e.g., Helvering v. Davis, 301 U.S. 619, 640-41 (1937) (finding that the Social Security program is a valid exercise of Congress’s spending power and is not unconstitutional under the Tenth Amendment); Steward Mach. Co. v. Davis, 301 U.S. 548, 589-91 (1937) (upholding the unemployment compensation provisions of the Social Security Act of 1935 as a valid use of the taxing power). 17 See Sunstein, supra note 3, 447-48 (noting that the “disintegrat[ion]” of challenges to executive agencies on nondelegation grounds led to a “working compromise in which broad delegations of power [to the executive branch] were tolerated”). 18 See, e.g., United States v. Carolene Prods. Co., 304 U.S. 144 (1938). 2017] The Myth of the Nondelegation Doctrine 383

scholars who labor on in the hope of a restoration, a second coming of the Constitution of liberty . . . .19 We argue that this familiar narrative is mistaken. There was no golden age in which the courts enforced a robust nondelegation doctrine that compelled legislators to make hard policy choices. The prevalent vision of the pre–New Deal nondelegation doctrine is a myth. Although the New Deal Court revised the formula for assessing whether Congress had improperly delegated legislative power to others, the change was more formal than real. The federal courts never posed a significant obstacle to the development of the administrative state and the delegation of extensive policymaking authority to executive officials. When it comes to the nondelegation doctrine, there is much less to the constitutional revolution of 1937 than is commonly thought. To appreciate the scope of the traditional nondelegation doctrine, one must cast a wide net. Even though the idea of the nondelegation doctrine was well established in pre–New Deal law and constitutional thought, the U.S. Supreme Court heard relatively few nondelegation cases prior to the New Deal. The real work of developing and applying nondelegation principles over the course of the nineteenth and early twentieth centuries was done by the state and lower federal courts. As a consequence, the pre–New Deal tradition of nondelegation jurisprudence has largely been left in obscurity. This Article seeks to recover that tradition and examine its contours. In doing so, it reveals that the constitutional limitation on the delegation of legislative power was frequently observed in theory but rarely enforced in practice. When it comes to the nondelegation doctrine, there is no “lost constitution” to recover. Our purpose in this Article is not to engage in the normative debate over whether the courts ought to enforce the nondelegation doctrine. Instead, our goal is descriptive in nature. We set out to uncover how the courts invoked the nondelegation doctrine in the period from the nation’s Founding through the New Deal. In order to accomplish this task, we compiled an original dataset of every federal and state case that involved a nondelegation challenge between 1789 and 1940. In Part I, we detail the conventional narrative surrounding the nondelegation doctrine. This narrative maintains that the doctrine was—throughout the nineteenth and early twentieth centuries—a restrictive limitation on legislative delegations but that it was ultimately cast into exile by the time of the New Deal constitutional revolution. In Part II, we examine how developments in three different areas have influenced the course of the nondelegation

19 Douglas H. Ginsburg, Delegation Running Riot, 1 REGULATION 83, 84 (1995) (reviewing DAVID SCHOENBROD, POWER WITHOUT RESPONSIBILITY: HOW CONGRESS ABUSES THE PEOPLE THROUGH DELEGATION (1993)). 384 University of Pennsylvania Law Review [Vol. 165: 379 doctrine. First, we explain the theoretical underpinnings of nondelegation constitutional principles and why they were thought to be central to American constitutionalism. Second, we examine the development of the nondelegation doctrine and its application by the U.S. Supreme Court, observing the Court’s repeated willingness to uphold congressional delegation of policymaking authority. Third, we turn to the states and the lower federal courts and identify a similar pattern of strong judicial statements about the importance of the nondelegation principle coupled with weak judicial enforcement of the doctrine. Finally, in Part III, we present empirical data on state and federal courts’ use of the nondelegation doctrine from the Founding through the New Deal.

I. THE “CONSTITUTION IN EXILE” It is generally believed that the classical Constitution of the nineteenth century included a nondelegation doctrine with real teeth, which was subsequently defanged as part of the struggle over the New Deal. The assertion and eventual abandonment of nondelegation principles are part of the conventional historical account of the constitutional transformation of the New Deal period. The tantalizing possibility that a robust nondelegation jurisprudence was “lost” in the New Deal revolution but could be recovered in the modern era to help constrain or dismantle the modern administrative state has been a recurring feature of libertarian constitutional thought since the Reagan era.20 In this Part, we recall those constitutional stories. Subsequent Parts will show that these stories are little more than myths. The significance of the nondelegation doctrine for the New Deal debates rests on the importance of the U.S. Supreme Court’s actions in Panama Refining Co. v. Ryan21 and A.L.A. Schechter Poultry Corp. v. United States.22 In Panama Refining Co., the Court struck down a provision of the National Industrial Recovery Act.23 Under the authority of that statute, the Roosevelt Administration had issued a code of “[f]air [c]ompetition” for the petroleum industry and began the process of setting production quotas for individual oil producers.24 Chief Justice Hughes objected that “Congress has declared no policy, has established no standard, has laid down no rule,” but merely delegated its lawmaking function to executive officials.25

20 E.g., Lawson, supra note 6. 21 293 U.S. 388 (1935). 22 295 U.S. 495 (1935). Chief Justice Hughes also thought the Bituminous Coal Conservation Act of 1935 relied on a theory that would “remove all restrictions upon the delegation of legislative power.” Carter v. Carter Coal Co., 298 U.S. 238, 318 (1936) (Hughes, C.J., concurring). 23 293 U.S. at 414-15, 433. 24 Id. at 408-10. 25 Id. at 430. 2017] The Myth of the Nondelegation Doctrine 385

In Schechter, the Court struck down a different provision of the same statute under which the Administration had created a “Live Poultry Code” to regulate the activities of those raising, buying, selling, and slaughtering chickens.26 Even Justice Cardozo was moved to declare that “[t]his is delegation running riot.”27 The Court’s emphasis on the idea that there must be some constitutional limits to how much policymaking authority could be delegated to the executive branch cast doubt upon the entire New Deal project.28 It was an idea that Assistant Attorney General Robert H. Jackson singled out as “purely judge-made, not Constitution-made.”29 Presumably, nondelegation was among those doctrines that a more “liberal-minded Judiciary” could expect to abandon without the need for a formal constitutional amendment.30 The demise of the nondelegation doctrine in the New Deal revolution is less famous than the contemporaneous doctrinal shifts in the interpretation of constitutional rights and federalism. No doubt the reconfiguration of the separation of powers is less prominent in part because the nondelegation doctrine had played a much smaller role in the long progressive lead-up to the final conflicts of the New Deal than had debates over substantive due process31 or interstate commerce.32 Moreover, the New Deal Court did not similarly repudiate the earlier nondelegation doctrine as it did in other areas of the law, and the turnaround on the nondelegation doctrine was neither as abrupt nor as dramatic. The constitutional revolution of 1937, contained snugly in volume 301 of the United States Reports, did not include a seminal case on the delegation of legislative powers. The reformation of the separation of powers came about somewhat more gradually and was not accompanied by sharp dissents lamenting the loss of ancient constitutional verities.

26 295 U.S. at 521-24. 27 Id. (Cardozo, J., concurring). 28 See, e.g., Daniel B. Rodriguez, The Administrative State and the Original Understanding: Comments on Eskridge and Ferejohn, 8 J.L. ECON. & ORG. 197, 200 (1992) (noting that “various New Deal and post–New Deal regulatory statutes . . . would potentially run afoul of a serious nondelegation doctrine”). 29 CARL BRENT SWISHER, AMERICAN CONSTITUTIONAL DEVELOPMENT 910 (1943); see also Patrick W. Duff & Horace E. Whiteside, Delegata Potestas Non Potest Delegari: A Maxim of American Constitutional Law, 14 CORNELL L.Q. 168, 195-96 (1929) (“Far from being a principle of constitutional law, it seems that our maxim has little, if any, application to the distribution of the work of government by the legislature. . . . The whole doctrine, insofar as it is asserted to be a principle of constitutional law, is built upon the thinnest of implication, or is the product of the unwritten super-constitution.”). 30 See President Franklin D. Roosevelt, A “Fireside Chat” Discussing the Plan for Reorganization of the Judiciary (Mar. 9, 1937) (announcing his plan to add additional Justices to the Supreme Court and thus make it more friendly to his New Deal proposals), in THE PUBLIC PAPERS AND ADDRESSES OF FRANKLIN D. ROOSEVELT, 122, 133 (Samuel I. Roseman ed., 1941). 31 See GILLMAN, supra note 10, at 2-5 (summarizing the body of scholarly work addressing the Court’s substantive due process jurisprudence in the Lochner era). 32 See BARRY CUSHMAN, RETHINKING THE NEW DEAL COURT: THE STRUCTURE OF A CONSTITUTIONAL REVOLUTION 139 (1998) (noting that the New Deal Commerce Clause cases “have been central to the ‘switch in time’ narrative”). 386 University of Pennsylvania Law Review [Vol. 165: 379

Nonetheless, contemporaries understood that the demise of the nondelegation doctrine was part and parcel of the reconstruction of the constitutional order that was wrought by the New Deal.33 Constitutional scholar and Roosevelt advisor Edward Corwin observed at the time that the “great structural principle of American constitutional government . . . supplied by the doctrine of the Separation of Powers” had been significantly revised “at the hands of the New Deal.”34 The Justices were simply coming to grips with a basic truth: “[T]hat the practice of delegated legislation is inevitably and inextricably involved with the whole idea of governmental intervention in the economic field.”35 “[I]n other words, governmental interventionism signifies the minimization of the principle of the Separation of Powers as a barrier preventing the fusion of presidential and congressional power.”36 Robert Cushman noted that when it came to the nondelegation doctrine, the Court seemed to prefer “eating its constitutional cake and having it too” by simultaneously upholding delegations of legislative power as “vitally necessary to the administration of government” and asserting that “legislative powers cannot be delegated.”37 The student editors of the Cornell Law Quarterly observed that though the courts “still pay lip service to the doctrine of the non-delegability of legislative powers, expansion of the operations of government has been accompanied by expansion of the limits of permissible delegation of legislative power to administrative bodies.”38 Reviewing legislative and judicial developments of the 1930s, one commentator thought the record “raise[s] the question as to whether the doctrine that delegated powers may not be redelegated has any significance in modern constitutional law.”39 The question thus raised could only be regarded as rhetorical. “[T]he maxim is no longer even substantially accurate.”40

33 An alternative narrative offered by some New Dealers is that the Court in Roosevelt’s first term had departed from established precedents and inappropriately imposed a new nondelegation requirement on Congress. See, e.g., EDWARD S. CORWIN, THE PRESIDENT: OFFICE AND POWERS, 1787–1984, 147 (Randall W. Bland et al. eds., 5th rev. ed. 1984) [hereinafter CORWIN, THE PRESIDENT] (summarizing pre–New Deal legislative delegation cases and arguing that “[i]n short, Congress may delegate its powers when it is necessary to do so in order to achieve the results it desires”). Others suggested that the industrial codes struck down by the Court were unique—and uniquely problematic—in involving the delegation of “governmental authority to private individuals.” Hugh Evander Willis, Constitution Making by the Supreme Court Since March 29, 1937, 15 IND. L.J. 179, 182 (1940). 34 CORWIN, CONSTITUTIONAL REVOLUTION, supra note 7, at 102. 35 Id. at 104. 36 Id. at 105. As Corwin later observed, the nondelegation doctrine had “been subsequently relegated by the Court to its increasingly crowded cabinet of juridical curiosities.” CORWIN, THE PRESIDENT, supra note 33, at 149. 37 Robert E. Cushman, The Constitutional Status of the Independent Regulatory Commissions, 24 CORNELL L.Q. 13, 27 (1938). 38 Jack L. Ratzkin, Notes and Comments, 26 CORNELL L.Q. 699, 699 (1941). 39 Charles B. Nutting, Congressional Delegations Since the Schechter Case, 14 MISS. L.J. 350, 367 (1942). 40 Id. 2017] The Myth of the Nondelegation Doctrine 387

Subsequent accounts of the New Deal constitutional revolution routinely take notice of the collapse of the nondelegation doctrine.41 During World War II, historian Benjamin Wright cautioned that because the Court had not formally overruled its early nondelegation cases, those principles remained a “shotgun-behind-the-door” that should temper congressional enthusiasm for excessive delegation.42 But after a few more years of experience under the New Deal regime, administrative law expert Kenneth Culp Davis advised lawyers simply that they would “do more harm than good to their clients’ interests” by appealing to the nondelegation doctrine in litigation; “[u]nrealistic verbiage in some of the older judicial opinions should not now be taken seriously.”43 Justice Thurgood Marshall declared that the nondelegation doctrine “is surely as moribund as the substantive due process approach of the same era—for which the Court is fond of writing an obituary . . . if not more so.”44 In recent decades, many conservative scholars and lawyers have called for a revival of the nondelegation doctrine that they see as having been cast aside in the constitutional revolution of the early twentieth century.45 Justice William Rehnquist took the lead, contending that the nondelegation cases of the 1930s “suffer from none of the excesses of judicial policymaking that plagued some of the other decisions of that era.”46 He chided his brethren, “We ought not to shy away from our judicial duty to invalidate unconstitutional delegations of legislative authority solely out of concern that we should thereby reinvigorate discredited constitutional doctrines of the pre–New Deal era.”47

41 See, e.g., SWISHER, supra note 29, at 964 (observing that whether due to better statutory drafting or “changes in the personnel of the Court,” no statutes were struck down on nondelegation grounds after 1937); see also WHITE, supra note 11, at 126 (portraying judicial resistance to executive agencies as part of what had to be overcome in the New Deal). 42 BENJAMIN F. WRIGHT, THE GROWTH OF AMERICAN CONSTITUTIONAL LAW 220 (1942). 43 1 KENNETH CULP DAVIS, ADMINISTRATIVE LAW AND GOVERNMENT 55 (1960). 44 Nat’l Cable Television Ass’n v. United States, 415 U.S. 352, 353 (1974). 45 See, e.g., Lawson, supra note 6, at 1240-41 (arguing that the modern administrative state is composed of “utterly vacuous statutes” that should be “easy kills under any plausible interpretation of the Constitution’s nondelegation principle”). The suggestion has led others in turn to raise cries that the conservative legal movement invited a return to “Black Monday,” when the Court struck down key components of the New Deal. See Alfred C. Aman, Jr., Introduction to Symposium: Bowsher v. Synar, 72 CORNELL L. REV. 421, 426 & n.29 (1987) (noting that some conservative Court decisions employed a formalist separation-of-powers approach reminiscent of Panama and Schecter); see also CASS R. SUNSTEIN, RADICALS IN ROBES: WHY EXTREME RIGHT-WING COURTS ARE WRONG FOR AMERICA 204-05 (2005) (“[L]arge-scale judicial revival of the nondelegation doctrine would do little to improve the operation of modern government. It might well make things worse, possibly much worse.”); Eric A. Posner & Adrian Vermeule, Interring the Nondelegation Doctrine, 69 U. CHI. L. REV. 1721, 1722 (2002) (“The nondelegation position lacks any foundation in constitutional text and structure, in standard originalist sources, or in sound economic and political theory.”). 46 Indus. Union Dep’t v. Am. Petroleum Inst., 448 U.S. 607, 675 (1980) (Rehnquist, J., concurring). 47 Id. at 686. 388 University of Pennsylvania Law Review [Vol. 165: 379

President Ronald Reagan’s Department of Justice suggested that the Court’s post–New Deal jurisprudence may be inconsistent with “a strict Madisonian concept of separation of powers.”48 Former Supreme Court-nominee Judge Douglas Ginsburg has admonished the Court for simply being too timid in the face of Congress to enforce a fundamental constitutional principle.49 Judge Ginsburg put his money where his mouth is in authoring a circuit court opinion attempting to revive the nondelegation doctrine.50 As amici, the libertarian Cato Institute has complained that the Court since the New Deal has “largely abdicated its responsibility” of enforcing the nondelegation doctrine.51 One Heritage Foundation writer aptly summed up a common view of modern conservatives: “[T]he administrative state is a profoundly unconstitutional form of government,” in part because of the necessary “delegation of legislative power” to executive agencies.52 And, as Gary Lawson complained, “The Supreme Court has not invalidated a congressional statute on nondelegation grounds since 1935. This has not been for lack of opportunity.”53

II. DEVELOPMENT OF THE NONDELEGATION DOCTRINE

A. The Constitutional Foundations The United States Constitution does not include an explicit provision recognizing the principle of separation of powers, but the commitment to some form of separation of powers across three branches of government is evident throughout the constitutional scheme and various provisions of the constitutional text. Article I of the Constitution specifies that “[a]ll legislative

48 OFFICE OF LEGAL POLICY, U.S. DEP’T OF JUSTICE, THE CONSTITUTION IN THE YEAR 2000: CHOICES AHEAD IN CONSTITUTIONAL INTERPRETATION 180 (1988). 49 See Douglas H. Ginsburg & Steven Menashi, Nondelegation and the Unitary Executive, 12 U. PA. J. CONST. L. 251, 264 (2010) (“[T]he judiciary, shrinking before the authority of the democratic legislature, has been complicit in allowing delegation to run riot.”). 50 See Am. Trucking Ass’ns v. EPA, 175 F.3d 1027, 1038 (D.C. Cir. 1999) (finding that the statutory language of the Clean Air Act, combined with an existing agency interpretation, constituted an unconstitutional delegation of legislative power). While the D.C. Circuit remanded the case for the agency to develop more appropriate, binding standards for its decisionmaking process, the case was ultimately reversed by Whitman v. Am. Trucking Ass’ns, 531 U.S. 457 (2001). 51 Brief for The Institute for Justice and The Cato Institute as Amici Curiae Supporting Respondents at 11, Whitman v. Am. Trucking Ass’ns, 531 U.S. 457 (2001) (No. 99-1257). 52 Joseph Postell, From Administrative State to Constitutional Government, HERITAGE FOUND. 4 (Dec. 7, 2012), http://www.heritage.org/research/reports/2012/12/from-administrative-state-to-const itutional-government [https://perma.cc/7S7J-YNDD]; see also Joel Hood, Before There Were Mouseholes: Resurrecting the Non-Delegation Doctrine, 30 BYU J. PUB. L. 123 (2015) (positing an originalist case that the Constitution requires a robust nondelegation doctrine); Lawson, supra note 6, at 1232 (arguing that “[t]he actual structure and operation of the national government today has virtually nothing to do with the Constitution”). 53 Lawson, supra note 6, at 1240 (footnote omitted). 2017] The Myth of the Nondelegation Doctrine 389

Powers herein granted shall be vested in a Congress of the United States.”54 Article II and Article III include their own vesting clauses, placing “[t]he executive Power” in the hands of the President of the United States55 and “[t]he judicial Power of the United States . . . in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.”56 The Constitution thus recognizes the existence of distinct functional powers that can be characterized as either legislative, executive, or judicial and places those powers in hands of different government entities or officials. To the extent that the Constitution departs from a pure separation-of-powers model and allows some sharing of powers across the branches of government, those exceptions are spelled out in the text.57 Whether to introduce some checks and balances into the constitutional system or to take advantage of some potential governmental efficiencies, the President is, for example, given a share of the legislative power through the prerogative of the presidential veto.58 Similarly, the Senate is given a share of the executive power through the right to advise and consent to the appointment of government officers.59 The sharing of political power across government branches means that the Constitution creates a certain “invitation to struggle” over the control of government policy.60 However, the Constitution also provides each branch of government a certain core of inalienable power and authority. There is no explicit textual prohibition on the delegation of legislative power to other actors, but such a rule has long been thought implicit in the U.S. Constitution. There are a variety of arguments explaining why a principle of nondelegation might be found in these textual provisions and the broader structure of the separation of powers. The very idea of a separation of powers might suggest that executive officials should refrain from, or be barred from, exercising legislative powers.61 Consolidating the legislative and executive functions in the same hands has long been seen as a serious threat to liberty, and a core principle of liberal constitutional theory was to separate those distinct governmental functions in distinct governmental organs. Montesquieu’s maxim that “[w]hen the legislative and executive powers are united in the same

54 U.S. CONST. art. I, § 1. 55 Id. art. II, § 1. 56 Id. art. III, § 1. 57 For a discussion of the differences between a pure separation-of-powers model and a checks-and-balances model of organizing the three branches, see M.J.C. VILE, CONSTITUTIONALISM AND THE SEPARATION OF POWERS 13-14 (2d ed. 1998). 58 U.S. CONST. art. I, § 7. 59 Id. art. II, § 2. 60 CORWIN, THE PRESIDENT, supra note 33, at 201. 61 Constitutional scholar M.J.C. Vile concluded that it was effectively impossible to reconcile the modern administrative state with traditional theories of the separation of powers among three distinct branches of government. VILE, supra note 57. 390 University of Pennsylvania Law Review [Vol. 165: 379 person, or in the same body of magistracy, there can be then no liberty”62 suggested not only that the constitutional Framers must be careful to separate the two classes of powers—but also that subsequent government officials must be prevented from concentrating the powers that had initially been set apart.63 Here the separation-of-powers concerns merge with due process concerns. Not only judges but average citizens should be able to understand what the rules are in order to be able to comply with them in systematic and predictable ways. If the command handed down by the legislature is murky, then judges are left to their own devices in determining how to apply them and citizens are at risk of being governed on an ad hoc basis by individual men rather than by settled law.64 Worse yet, judges are left with no standard by which to evaluate the actions of executive officials. There is little effective difference between the exercise of power under an expansive delegation of power and the simple exercise of arbitrary discretion. In neither case could a third party, such as a judge, assess whether the action taken by a government official was authorized or constrained by a preexisting rule. Robert Cushman went so far as to argue, “[T]he doctrine of the non-delegability of legislative power could safely be scrapped as long as due process of law remains the effective constitutional guarantee it now is.”65 Due process considerations and nondelegation considerations would generate the same set of constitutional principles. Somewhat differently, some have argued that the lawmaking function should reside specifically in a representative assembly, which suggests that it would be inappropriate to transfer that power to a less accountable and less representative institution.66 The English political theorist John Locke had this in mind in contending that the legislative cannot transfer the power of making laws to any other hands; for it being but a delegated power from the people, they who have it cannot

62 BARON DE MONTESQUIEU, THE SPIRIT OF THE LAWS 202 (David Wallace Carrithers ed., Thomas Nugent trans., 1977) (1748). 63 A closely related argument would emphasize due process or rule-of-law principles. Combining legislative and executive power risks subjecting individuals to arbitrary or purely discretionary power rather than to regularized and known rules. See, e.g., Cushman, supra note 37, at 23-24 (arguing that the Supreme Court’s due process is a “flexibile and practicable doctrine” to respond to regulation that attempts to fuse power across the three branches). 64 Delegation can thus be understood as at odds with the Benthamite desire for legislative codification and for legal transparency that had once accompanied the rise of legislatures as significant policymaking bodies.See, e.g., CECIL T. CARR, DELEGATED LEGISLATION 1 (1921) (“The action of our Acts of Parliament grows more and more dependent upon subsidiary legislation. More than half our modern Acts are to this extent incomplete statements of law.”). 65 Cushman, supra note 37, at 33. 66 David Schoenbrod makes a related argument, contending that expansive delegation of lawmaking authority to administrative agencies simply results in worse public policy. DAVID SCHOENBROD, POWER WITHOUT RESPONSIBILITY: HOW CONGRESS ABUSES THE PEOPLE THROUGH DELEGATION 125-31 (1993). 2017] The Myth of the Nondelegation Doctrine 391

pass it over to others. . . . And when the people have said, we will submit to rules, and be governed by laws made by such men, and in such forms, nobody else can say other men shall make laws for them; nor can the people be bound by any laws but such as are enacted by those whom they have chosen and authorized to make laws for them.67 The late-nineteenth century jurist Thomas Cooley referenced this principle when arguing that “[t]his high prerogative has been intrusted to [the legislature’s] own wisdom, judgment, and patriotism, and not to those of other persons.”68 The Lockean postulate that the legislature cannot transfer its lawmaking power to another can also be understood in more formalistic terms. Locke pointed out that “[t]he people alone can appoint the form of the commonwealth,” and thus mere government officials should not be able to alter those forms without the consent of the governed.69 The “maxim of the common law[,] [d]elegata potestas non potest delegari,” is frequently cited as authority for the general point that those who are entrusted with power must exercise the trust themselves and not further delegate the power “to a stranger, whose ability and integrity might not be known to the principal, or, if known, might not be selected by him for such a purpose.”70 Sotirios Barber has characterized this as a simple matter of “constitutional supremacy.”71 Once the “constituent act of establishing government [is complete . . .] neither the government nor any of its parts should change the constitutional arrangement of offices and powers.”72 Congress may not “abdicate” the legislative power that was entrusted to it alone by the constitutional Framers. Of course, there have been many skeptics. Justice Robert Jackson thought it “perfectly obvious” that the Constitution and its Framers “contemplated a large measure of delegation” and would not have thrown up obstacles to such a necessary measure of governance.73 In his influential administrative law casebook, Walter Gellhorn complained that the courts had shown a “lack of logic” in their nondelegation cases, leading them to repeatedly yield to “necessity”

67 JOHN LOCKE, SECOND TREATISE OF GOVERNMENT § 141 (Thomas P. Peardon ed., Liberal Arts Press 1952) (1690). 68 THOMAS M. COOLEY, THE GENERAL PRINCIPLES OF CONSTITUTIONAL LAW IN THE UNITED STATES OF AMERICA 97 (1880). In characterizing the mid-twentieth century American state, Theodore Lowi linked a concern with democracy with a concern for the rule of law. See THEODORE J. LOWI, THE END OF LIBERALISM 125 (2d ed. 1979) (“A government of statutes without standards may produce pluralism, but it is a pluralism of privilege and tight access . . . .”). 69 LOCKE, supra note 67, § 141. 70 JOSEPH STORY, COMMENTARIES ON THE LAW OF AGENCY § 13 (1839). 71 SOTIRIOS A. BARBER, THE CONSTITUTION AND THE DELEGATION OF CONGRESSIONAL POWER 37 (1975). 72 Id. 73 ROBERT H. JACKSON, THE STRUGGLE FOR JUDICIAL SUPREMACY 92-93 (1941). 392 University of Pennsylvania Law Review [Vol. 165: 379 while continuing to give lip service to an unworkable rule.74 Louis Jaffe declared simply that “[d]elegation of ‘lawmaking’ power is the dynamo of modern government,” and the legal objection to it is little more than the carping of the “holders of economic power.”75 The political scientist John Roche concluded that the original conception of the separation of powers established only a “fairly simple and matter-of-fact division of the agencies of the national government” and posed no bar to the expansive delegation of legislative powers to the executive.76 The putative prohibition on delegations of legislative power “has been demolished by constitutional logic drawn from John Marshall” holding that congressional power was plenary and that any limits on the delegation of power should be understood as “political, not constitutional, in character.”77 Since the New Deal, the skeptics have been predominant.78

B. The Doctrine in the Supreme Court The Supreme Court periodically grappled with the nondelegation doctrine over the course of its history up to and through the New Deal.79 Early in its history, the Court recognized the constitutional significance of the principles that implied limits on the authority of Congress to delegate its lawmaking power to other government officials. The Court was periodically called upon to elaborate on those principles, apply them to new contexts, and evaluate new legislative innovations in how Congress has sought to set public policy. The Court’s own track record prior to the New Deal was one of uniform deference to congressional decisions to delegate some rulemaking to others. While the Court briefly deployed the same doctrines to veto congressional legislation during the New Deal, it quickly retreated to its earlier deferential posture. The Court’s own history suggests that the battles of the New Deal should be viewed as an idiosyncratic departure from the constitutional norm rather than representative of how the Constitution had historically been understood and applied. The Court first addressed a challenge to a federal statute as making an unconstitutional delegation of legislative power during the Jefferson

74 WALTER GELLHORN, ADMINISTRATIVE LAW: CASES AND COMMENTS 175 (1940). 75 Louis L. Jaffe, An Essay on Delegation of Legislative Power: I, 47 COLUM. L. REV. 359, 359 (1947). 76 John P. Roche, Distribution of Powers, in 3 INTERNATIONAL ENCYCLOPEDIA OF THE SOCIAL SCIENCES 300, 305-07 (David L. Sills ed., 1968). 77 Id. at 306-07. 78 See, e.g., Peter H. Schuck, Delegation and Democracy: Comments on David Schoenbrod, 20 CARDOZO L. REV. 775, 778 (1999) (“[W]e may need more delegation to agencies, not less.”). 79 For a detailed look at the key cases, see generally Andrew J. Ziaja, Hot Oil and Hot Air: The Development of the Nondelegation Doctrine Through the New Deal, a History, 1813–1944, 35 HASTINGS CONST. L.Q. 921 (2008). 2017] The Myth of the Nondelegation Doctrine 393

Administration.80 The Non-Intercourse Act of 1809 replaced the highly controversial trade embargo that was designed to prevent the United States from becoming entangled in the war between Britain and France by keeping American ships at home.81 While the earlier legislation imposed a comprehensive trade embargo, the Non-Intercourse Act directed the embargo specifically against the warring powers of Britain and France and also authorized the President to lift the embargo against either country if he were to recognize the neutral commerce rights of the United States.82 The terms of the Act were extended by Congress in 1810, while Congress also temporarily suspended the implementation of the Act, giving the two European nations a fixed period to renounce their wartime policies against American shipping.83 The President was charged with the responsibility of announcing whether Britain or France had come to terms with the United States.84 Britain did not give in, and the terms of the embargo were put into effect.85 Customs officials subsequently seized the cargo of the brig Aurora in the port of New Orleans for violating the embargo against Great Britain by importing goods from Liverpool.86 The owner of the cargo brought suit in federal court, arguing (among other issues) that the terms of the Non-Intercourse Act of 1810 were unconstitutional.87 In particular, the owner argued that Congress had impermissibly “transfer[red] the legislative power to the President” and gave a presidential proclamation “the force of a law.”88 As the lawyer for the Aurora argued: “Whoever heard of a conditional penal law . . . ?”89 Unusually, Chief Justice John Marshall did not write the opinion of the Court in the case but instead left those duties to the Jeffersonian Justice William Johnson.90 Johnson did not bother to examine the constitutional question in any detail. He simply observed that the Court could “see no sufficient reason, why the legislature should not exercise its discretion in reviving the act . . . either expressly or conditionally, as their judgment should direct.”91 Although the nondelegation issue was raised by counsel, the Justices themselves did not elaborate any particular principles for evaluating such

80 For a review of the Court’s early cases addressing legislative delegation, see Keith E. Whittington, Judicial Review of Congress Before the Civil War, 97 GEO. L.J. 1257, 1291-92 (2009). 81 Cargo of the Brig Aurora v. United States, 11 U.S. (7 Cranch) 382, 382-85 (1813). 82 Id. at 383. 83 Id. at 383-84. 84 Id. at 384. 85 Id. 86 Id. at 382. 87 Id. 88 Id. at 386. 89 Id. 90 Id. at 387. 91 Id. at 388. 394 University of Pennsylvania Law Review [Vol. 165: 379 situations and largely dismissed the framing of the law as one of delegating “legislative power to the President.”92 Instead, the Court left the President’s statutorily specified role in triggering the trade embargo to the side and focused on the legislature’s power to exercise its own discretion to extend the embargo conditionally “upon the occurrence of any subsequent combination of events.”93 The active agency of the President was downplayed, while the factual preconditions for legal action were underscored. Implicitly, the framing of the opinion suggested that the President acted simply as a fact-finder, not as a lawmaker. The President “revived” the statute only in a mechanical sense of making known whether the factual conditions specified by Congress had been met. As a result, the Marshall Court gave its approval to the possibility of conditional legislation, while saying little on the potential problem of excessive delegation of lawmaking power.94 In the 1820s, the Court considered constitutional challenges to two other statutes that required it to address the nondelegation principle more directly. The first case was Wayman v. Southard, a case involving two related issues of delegation that arose from the compromised quality of the Judiciary Act of 1789.95 The statute required that “the laws of the several states,” except where otherwise provided, “shall be regarded as rules of decision in trials at common law in the courts of the United States.”96 Rather than creating a distinct system of federal rules of judicial procedure, Congress preferred that the federal courts be in harmony with the states in which they sat. The easiest way to do that—especially since state law was constantly evolving—was simply to piggyback on judicial procedures that the states had already put in place. In effect, Congress had delegated the development of federal civil procedure to the states. Moreover, such rules were often going to be developed not by legislatures but by judges.

92 Id. at 386. 93 Id. at 388. 94 The Court likewise left the nondelegation issues somewhat implicit when resolving a constitutional challenge to congressional antipiracy statutes. See United States v. Smith, 18 U.S. (5 Wheat.) 153 (1820). Daniel Webster argued that Congress itself was obliged to define the crime of piracy and could not delegate to the courts the power to define such crimes. See id. at 156-57 (“Congress is bound to define it, in terms, and is not at liberty to leave it to be ascertained by judicial interpretation.”). Justice Joseph Story responded that piracy was “a term of a known and determinate meaning, as by an express enumeration of all the particulars included in that term.” Id. at 159. Judges could be understood as interpreting and applying, rather than legislating, when determining what acts constituted the crime of piracy. See id. The issue recurs in later cases, with the same result. See, e.g., In re Kollock, 165 U.S. 526, 533 (1897) (“The criminal offence is fully and completely defined by the act and the designation by the Commissioner of the particular marks and brands to be used was a mere matter of detail.”). 95 23 U.S. (10 Wheat.) 1 (1825). 96 Judiciary Act of 1789, ch. 20, § 34, 1 Stat. 73, 92 (current version at 28 U.S.C. § 1652 (2006)). 2017] The Myth of the Nondelegation Doctrine 395

The attorney before the Court in Wayman emphasized that substantive interests were at stake in regulating judicial procedures, such as the remedies available on violated contracts.97 He contended that the “power of making such regulations is exclusively vested in the legislative department . . . . It is the office of the legislator to prescribe the rule, and of the Judge to apply it.”98 In response, Chief Justice Marshall admitted that Congress could not “delegate to the Courts, or to any other tribunals, powers which are strictly and exclusively legislative.”99 The question was what powers might not be “strictly and exclusively legislative,” such that Congress might choose not to exercise them itself.100 Marshall acknowledged that [t]he line has not been exactly drawn which separates those important subjects, which must be entirely regulated by the legislature itself, from those of less interest, in which a general provision may be made, and power given to those who are to act under such general provisions to fill up the details.101 Marshall was not prepared to engage in the “delicate and difficult inquiry” that would be necessary to draw a “precise boundary” between permissible and impermissible delegations.102 It was enough to recognize that “the maker of the law may commit something to the discretion of the other departments” and that this particular delegation was well-considered to advance the national interests of the time.103 Justice Smith Thompson addressed the same issue a few years later in Bank of the United States v. Halstead.104 He observed that regulation of the judicial process merely related “to the ministerial duty” of court officers and “partakes no more of legislative power, than that discretionary authority intrusted to every department of the government in a variety of cases.”105 But Thompson also seemed to dodge what Justice Marshall characterized as the “delicate and difficult” inquiry of drawing lines between appropriate and

97 Wayman, 23 U.S. (10 Wheat.) at 13-14. 98 Id. at 14. 99 Id. at 42-43. 100 Id. 101 Id. at 43. 102 Id. at 46. 103 Id. at 45-46. The Marshall Court obscured the difference between delegating judicial procedures to federal judges or to state legislatures and judges. The Court later addressed the problem of federal delegation to the states more directly when considering a federal statute authorizing state legislatures to establish mining regulations on federal lands within their states. It was not, the Court stated, “of a legislative character in the highest sense of the term” to establish such regulations but merely a “determination of minor matters,” while again observing that the federal statutes in question had been in place for many years and substantial reliance interest had been built up around their presumed validity. Butte City Water Co. v. Baker, 196 U.S. 119, 126 (1905). 104 23 U.S. (10 Wheat.) 51, 61-62 (1825). 105 Id. 396 University of Pennsylvania Law Review [Vol. 165: 379 inappropriate delegations by leaning on a long history of usage or practice.106 If “any doubt existed” about the constitutionality of the delegation in such cases, “the practical construction heretofore given to it, ought to have great weight.”107 According to Justice Thompson, the courts had been operating without incident under this system of rules for nearly thirty years and the time to ask whether Congress was constitutionally authorized to set up that system had passed.108 The Court had remarkably little to say regarding the delegation of legislative power from the late Marshall Court through the remainder of the nineteenth century.109 While the Marshall Court largely avoided serious engagement with the principles and standards of nondelegation, the Court finally and influentially addressed the question directly at the end of the nineteenth century. In the Gilded Age, both the federal and state governments began to experiment with new regulatory schemes and governmental institutions which blurred the traditional boundaries between legislatures, executives, and courts.110 But it was not an example of the growing administrative state that generated the first significant federal case grappling with the problem of excessive legislative delegation. Instead, like the earliest nondelegation case, the case of Field v. Clark involved the regulation of international trade and presidential participation in triggering statutory application.111 The Tariff Act of 1890 authorized the President, by proclamation, to trigger a higher duty rate on specified goods from individual countries that did not engage in reciprocal free trade with the United States.112 As the Marshall Court had done, the Field Court observed that “Congress has frequently, from the organization of the government to the present time, conferred upon the President powers, with

106 See id. at 62-63. 107 Id. The Court later articulated a distinct constraint on the ability of Congress to delegate legislative powers to the courts: Article III courts could not be tasked to perform duties that were not encompassed by the cases and controversies requirement. See Keller v. Potomac Elec. Power Co., 261 U.S. 428, 441-43 (1923) (holding that Congress could not authorize a federal district court to set utility rates in the District of Columbia because that would be an exercise of legislative power, not judicial power). 108 Halstead, 23 U.S. (10 Wheat.) at 63-64. 109 The Court briefly recognized, as the state courts had, that it was an appropriate public purpose for state legislatures to delegate to railroad corporations the state’s power of eminent domain. See, e.g., Queensbury v. Culver, 86 U.S. (19 Wall.) 83, 90-91 (1873) (upholding a state law authorizing towns to issue bonds that helped railroads acquire land); Olcott v. Supervisors, 83 U.S. (16 Wall.) 678, 691 (1872) (“[S]uch delegation of power can be justified . . . [if] the property taken by these [railroad] companies is taken for the public use.”). 110 See Herbert Hovenkamp, Regulatory Conflict in the Gilded Age: Federalism and the Railroad Problem, 97 YALE L.J. 1017, 1017 (1988) (describing the Gilded Age as a period in which “the economics and the politics of business regulation were going through a period of convulsive change”). 111 143 U.S. 649, 692-93 (1892). 112 Id. at 680. 2017] The Myth of the Nondelegation Doctrine 397 reference to trade and commerce” and noted that this “fact is entitled to great weight in determining the question before us.”113 But the Field Court was more emphatic than the Marshall Court about the signficance of the underlying principle: that “Congress cannot delegate legislative power to the President is a principle universally recognized as vital to the integrity and maintenance of the system of government ordained by the Constitution.”114 And the Court was more willing to try to draw the line between permissible and impermissible delegation.115 Justice Harlan explained, [The challenged portion of the Tariff Act of 1890] is not inconsistent with that principle. It does not, in any real sense, invest the President with the power of legislation. . . . Congress itself determined that the provision of the [A]ct . . . should be suspended as to any country producing and exporting [certain goods], that imposed exactions and duties on the agricultural and other products of the United States, which the President deemed, that is, which he found to be, reciprocally unequal and unreasonable. Congress itself prescribed in advance the duties to be levied, collected and paid . . . . Nothing involving the expediency or the just operation of such legislation was left to the determination of the President.”116 The President had “no discretion,” but merely “ascertained the existence of a particular fact” that Congress had specified as necessary to trigger certain statutory features.117 “Legislative power was exercised when Congress declared that the suspension should take effect upon a named contingency.”118 The President was not exercising legislative will and “making law[],” but was a “mere agent of the law-making department to ascertain and declare the event upon which its expressed will was to take effect.”119 Citing earlier state court decisions, the Court put the emphasis on the locus of “discretion.”120 Whatever governmental agency was exercising discretion was exercising the lawmaking function: Congress was not delegating lawmaking power if it simply specified a series of conditions under which different statutory provisions would come into effect and designated an agent to determine whether those conditions had been met.121 The Court did address the more standard problem of the administrative state a few years later, but obscured the challenge that it posed to how the

113 Id. at 683. 114 Id. at 692. 115 Id. at 692-93. 116 Id. 117 Id. at 693. 118 Id. 119 Id. 120 Id. at 693-94. 121 Id. 398 University of Pennsylvania Law Review [Vol. 165: 379

Justices had previously discussed the nondelegation principle. The Court backed into the issue through a Fourteenth Amendment challenge to a state mine regulation.122 The statute simply directed that an inspector inspect each mine “as often as he may deem it necessary and proper.”123 The question before the Court was whether it was consistent with the requirements of constitutional due process to give executive officials discretion to classify mines and determine a schedule of inspections.124 Absent “abuse of . . . discretion” by the inspectors, the Court was unconcerned that they were entrusted with discretionary authority.125 In justifying this position, the Court recognized that while “legislative power cannot be delegated,” there are “some exceptions to the rule” in which Congress can create conditional statutes that utilize presidential discretion—as in Aurora and Field; the Court then pivoted without explanation to the present situation, finding the mining regulation to be a reasonable exception to the nondelegation rule.126 The Court held that “in case the legislature find[s] it impracticable to classify the mines for the purposes of inspection, [it can] commit that power to a body of experts who are not only experienced in the operation of mines, but are acquainted with the details necessary to be known to make a reasonable classification.”127 Although framed in the context of the Fourteenth Amendment, the Court’s claim about nondelegation was a general one. It simply seemed “obviously necessary” that something as fact-specific as how often a particular mine needed to be inspected be determined “by some executive officer” with the requisite “practical knowledge” to make such decisions.128 Shortly afterwards, in a case involving the Secretary of Treasury’s discretionary authority to exclude adulterated tea from the American marketplace, the Court married the reasoning of Field to this new concern about the expert knowledge of executive officials.129 In Buttfield v. Stranahan, the Court referenced Field to claim that “Congress legislated on the subject as far as was reasonably practicable” and that denying Congress the power to invest executive officials with discretionary werpo would functionally declare “that the plenary power vested in Congress to regulate foreign commerce could not be efficaciously exerted.”130 The delegation of discretionary authority to

122 St. Louis Consol. Coal Co. v. Illinois, 185 U.S. 203, 204-06 (1902). 123 Id. at 208 (emphasis in original). 124 Id. at 207-09. 125 Id. at 209-10. 126 Id. at 210. 127 Id. at 211. 128 Id.; see also Douglas v. Noble, 261 U.S. 165, 169-70 (1923) (holding that dentistry licensing standards could be delegated to an administrative board). 129 See Buttfield v. Stranahan, 192 U.S. 470 (1904). 130 Id. at 496. 2017] The Myth of the Nondelegation Doctrine 399 executive officers was simply a necessary and proper means for exercising an enumerated power. An executive official’s use of discretionary power “was committed to his judgment, to be honestly exercised,” and the Court declined to decide whether the executive official’s standards may have been different from Congress’s intent.131 From there, the Court was open to Congress deciding whether to investigate and address if individual bridges obstructed interstate commerce and then legislate regarding each bridge or declare “a general rule and impose[] upon the Secretary of War the duty of ascertaining what particular cases came within the rule prescribed by Congress.”132 So long as the Secretary of War “will only execute the clearly expressed will of Congress,” then “[h]e could not be said to exercise strictly legislative or judicial power.”133 It was “impracticable in view of the vast and varied interests which require National legislation from time to time” for Congress to occupy itself with investigating individual obstructions to interstate commerce.134 Simply stating that “navigation should be freed from unreasonable obstructions” was sufficient guidance to executive officials.135 If Congress could not take such a step, it would “stop the wheels of government” and “bring about confusion, if not paralysis, in the conduct of the public business.”136 In a subsequent case, the Court went even further and asserted that executive officers “did not legislate” so long as they “did not go outside of the circle of that which the act itself had affirmatively required to be done, or treated as unlawful if done.”137 If government officials were “confining themselves within the field covered by the statute,” it did not matter that there was no apparent expressed will of Congress regarding the specific rules to be issued by the executive branch.138 By the Progressive Era, the Court was willing to characterize almost any action that a government official performed as nonlegislative. When the Interstate Commerce Commission required all of its regulated businesses to adopt a uniform system of accounting, the Court said that this was simply “the carrying out of details” regarding a statutory directive that required regulated businesses to

131 Id. at 496-97. 132 Union Bridge Co. v. United States, 204 U.S. 364, 386 (1907). 133 Id. 134 Id. 135 Id. 136 Id. at 387. The problem of the “generic statutory provision” was potentially even greater in the case of the Sherman Antitrust Act, but the Court was quick to say that judges were doing nothing more than performing the “duties which that department of the government has exerted from the beginning” and simply determining “whether a particular act or acts are within a given prohibition.” Standard Oil Co. of N.J. v. United States, 221 U.S. 1, 69-70 (1911). 137 United States v. Grimaud, 220 U.S. 506, 518 (1911). 138 Id. 400 University of Pennsylvania Law Review [Vol. 165: 379 supply an annual report to the Commission.139 It was sufficient that Congress “laid down general rules for the guidance of the Commission.”140 In another case, the Court examined whether delegating to a board the power to censor films that were not “educational, moral, amusing or harmless” violated the nondelegation principle by failing to provide any protection against the “arbitrary judgment, whim and caprice” of the censor.141 The Court attempted to assuage this concern by assuring the film industry that such “general terms[] get precision from the sense and experience of men.”142 The Court reasoned that if legislatures were required to be more specific when empowering government officials, “the many administrative agencies created by the state and [n]ational governments would be denuded of their utility and government in some of its most important exercises [would] become impossible.”143 Similarly, in Mahler v. Eby, the Court held that, although the “executive may not exercise [the power to expel aliens] without congressional authority,” Congress may delegate the authority to deport “undesirable residents of the United States.”144 “[T]he expression ‘undesirable residents of the United States’ is sufficiently definite” to delegate this authority and guide the executive branch.145 After all, our “history has created a common understanding of the words ‘undesirable residents’ which gives them the quality of a recognized standard.”146 In a different case, Congress had delegated to the Federal Radio Commission sweeping authority to allocate radio frequencies.147 The Court held that if the courts interpreted the guidelines that licenses be granted “as public convenience, interest, or necessity requires” as not “so indefinite as to confer an unlimited power,” then the Commission would be implementing the will of Congress rather than exercising legislative discretion.148 Furthermore, in Wisconsin v. Illinois, the Court assured the states bordering Lake Michigan that the Secretary of War was not exercising legislative authority when he issued permits allowing the construction of canals that lowered the level of the lake because such matters were “a peculiarly expert question” and thus “naturally within the executive function.”149 All of these cases suggest an underlying principle that restrains the nondelegation doctrine. Specifically, if something

139 Interstate Commerce Comm’n v. Goodrich Transit Co., 224 U.S. 194, 215 (1912). 140 Id. 141 Mut. Film Corp. v. Indus. Comm’n of Ohio, 236 U.S. 230, 245 (1915). 142 Id. at 245-46. 143 Id. at 246. 144 264 U.S. 32, 40 (1924). 145 Id. 146 Id. 147 Fed. Radio Comm’n v. Nelson Bros. Bond & Mortg. Co., 289 U.S. 266, 276 (1933). 148 Id. at 285. 149 278 U.S. 367, 414 (1929). 2017] The Myth of the Nondelegation Doctrine 401 seems “impracticable for Congress” to do, then it can be delegated as an “administrative function[].”150 Chief Justice William Howard Taft’s opinion in J.W. Hampton, Jr. & Co. v. United States is notable for attempting to propound a clearer and more robust doctrine for evaluating legislative delegations.151 In the Tariff Act of 1922, Congress created a “flexible tariff provision” that authorized the President to adjust tariff duties on imported goods to erase any gap between the production costs of those goods in the United States and competing foreign countries.152 The Court reasoned that it would be practically difficult for Congress to determine by itself differences in relative costs and make timely adjustments to tariffs in response; thus, delegating the authority to make such determinations and set duties accordingly was reasonably delegated to a tariff commission.153 In justifying this outcome, Chief Justice Taft offered a more elaborate statement of the nondelegation principle than the Court had previously done. Quoting the “well-known maxim ‘Delegata potestas non potest delegari,’”154 Taft stated it would be “a breach of the National fundamental law if Congress gives up its legislative power and transfers it to the President.”155 Congress was free to “seek[] assistance” and “invoke the action” of the other branches so long as “the action invoked shall not be an assumption of the constitutional field of action of another branch.”156 But determining whether the legislature had gone too far was a matter of “common sense and the inherent necessities of the governmental co-ordination.”157 So long as “Congress shall lay down by legislative act an intelligible principle to which the person or body authorized to fix such rates is directed to conform, such legislative action is not a forbidden delegation of legislative power.”158 Executive officials did not exercise the “power of legislation” if “nothing involving the expediency or just operation of such legislation was left to [their] determination.”159 In such circumstances, the executive was “the mere agent of the law-making department,” no matter how much discretion might be allowed when exercising that agency.160 The “intelligible

150 See United States v. Shreveport Grain & Elevator Co., 287 U.S. 77, 85 (1932) (approving the power of the Food and Drug Administration to determine “reasonable variations” from the requirement that the quantity of goods in a package be “plainly and conspicuously marked”). 151 276 U.S. 394, 400, 404-05 (1928). 152 Id. at 400; see also Tariff Act of 1922, ch. 356, 42 Stat. 858 (codified as amended at 19 U.S.C. § 1526 (2012)) (requiring the President to levy duties whenever such a gap existed). 153 Hampton, 276 U.S. at 404-05. 154 Id. at 405. 155 Id. at 406. 156 Id. 157 Id. 158 Id. at 409. 159 Id. at 410. 160 Id. at 411. 402 University of Pennsylvania Law Review [Vol. 165: 379 principle” standard remains the Court’s primary vehicle for assessing whether an impermissible delegation of legislative power has taken place.161 Prior to 1935 the Court had never struck down a federal statutory provision as an unconstitutional delegation of power to the executive. When it did so in considering provisions of the National Industrial Recovery Act of 1933, the Court purported to apply the same standards that it had traditionally used in nondelegation cases.162 As the Court construed it, the provision at issue in Panama Refining “gives to the President an unlimited authority to determine the policy and to lay down the prohibition, or not to lay it down, as he may see fit.”163 In sum, “Congress left the matter to the President without standard or rule, to be dealt with as he pleased.”164 Congress had authorized action, but had not set forth any policy regarding what sort of action ought to be taken and in what circumstances. But even in striking down the “hot oil” provision, the Court emphasized, [L]egislation must often be adapted to complex conditions involving a host of details with which the national legislature cannot deal directly. The Constitution has never been regarded as denying to the Congress the necessary resources of flexibility and practicality, which will enable it to perform its function in laying down policies and establishing standards, while leaving to selected instrumentalities the making of subordinate rules within prescribed limits.165 Nonetheless, the Court had always said there were some limits to the ability of Congress to delegate rulemaking authority to others.166 The Court thought the early New Deal statutes were unique in establishing “no requirement, no definition of circumstances and conditions in which” the President should or should not act.167 Congress had not made a policy; it had delegated to the President the authority to make a policy regarding a specified

161 See, e.g., Mistretta v. United States, 488 U.S. 361, 372-73 (1989) (describing the use of the intelligible principle standard in a case challenging federal sentencing guidelines). 162 See Panama Ref. Co. v. Ryan, 293 U.S. 388, 430 (1935) (observing that “the Court has recognized that there are limits of delegation which there is no constitutional authority to transcend . . . [and where] Congress has declared no policy, has established no standard, [and] has laid down no rule,” the statute goes beyond the permissible scope of delegation); see also A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 541 (1935) (finding that a provision is unprecedented given prior delegation decisions because it provides no rules of conduct or standards for operating). The Court’s opinion in Carter v. Carter Coal Co. is more ambiguous on this point. The majority opinion takes up a “legislative delegation in its most obnoxious form,” but the concern is with entrusting to a private group “the power to regulate the affairs of an unwilling minority,” with the exercise of coercive government power by private actors rather than the exercise of a lawmaking authority. 298 U.S. 238, 311 (1936). 163 Panama Ref., 293 U.S. at 415. 164 Id. at 418. 165 Id. at 421. 166 Id. at 430. 167 Id. 2017] The Myth of the Nondelegation Doctrine 403 subject matter. Congress had done what the Court had previously said it could not—it had granted an executive official “unlimited power” over a particular subject matter.168 Given this history of purported doctrinal continuity, it is perhaps not surprising that the Court did not repudiate the nondelegation arguments in the 1935 cases. After the 1937 “switch in time,” the Court continued to address nondelegation claims in generally the same terms as it had before the New Deal. The Court, for example, distinguished the Tobacco Inspection Act of 1935 from the National Industrial Recovery Act, claiming this “is not a case where Congress has attempted to abdicate, or to transfer to others, the essential legislative functions with which it is vested by the Constitution.”169 The Court was back to emphasizing that “legislation must often be adapted to conditions involving details with which it is impracticable for the legislature to deal directly.”170 In the Tobacco Inspection Act, the directive that the Secretary of Agriculture “establish standards for tobacco by which its type, grade, size, condition, or other characteristics may be determined”171 was of a kind with “familiar legislative practice” and was sufficient for Congress to “set forth its policy” for the executive to follow.172 Similarly, the Court found Congress appropriately delegated its authority in the Agricultural Marketing Agreement Act of 1937 by stating “the purpose which the Congress seeks to accomplish and the standards by which that purpose is to be worked out with sufficient exactness to enable those affected to understand these limits.”173 The Court noted, “Congress needs specify only so far as is reasonably practicable.”174 With the Agricultural Adjustment Act of 1938, Congress delineated the “considerations which are to be held in view” when the Secretary of Agriculture sets production quotas and also provided subsequent administrative and judicial review “to correct errors,” which was sufficient to “protect against arbitrary action” by an executive official.175 Regarding the Bituminous Coal Conservation Act of 1935, the Court concluded that, “in the hands of experts[,] the criteria which Congress has supplied [for fixing coal prices] are wholly adequate for carrying out the general policy and purpose of the Act.”176 The Court again warned that “if Congress were under the constitutional compulsion of filling in the details” of

168 See Fed. Radio Comm’n v. Nelson Bros. Bond & Mortg. Co., 289 U.S. 266, 285 (1933). 169 Currin v. Wallace, 306 U.S. 1, 15 (1939). 170 Id. 171 Tobacco Inspection Act, 49 Stat. 731, 732 (1935). 172 Currin, 306 U.S. at 16-17. 173 United States v. Rock Royal Coop., Inc., 307 U.S. 533, 574 (1939). 174 Id. 175 Mulford v. Smith, 307 U.S. 38, 49 (1939). 176 Sunshine Anthracite Coal Co. v. Adkins, 310 U.S. 381, 398 (1940). 404 University of Pennsylvania Law Review [Vol. 165: 379 the policy, then “the burdens of minutiae would be apt to clog the administration of the law and deprive the agency of that flexibility and dispatch which are its salient virtues.”177 When confronted with a nondelegation challenge to the wartime Emergency Price Control Act of 1942, the Court again highlighted the differences between the “exercise by Congress of its legislative power” represented by this statute and the National Industrial Recovery Act (NIRA) struck down in Schechter.178 In the wartime statute, Congress “has stated the legislative objective, has prescribed the method of achieving that objective—maximum price fixing—, and has laid down standards to guide the administrative determination of both the occasions for the exercise of the price-fixing power, and the particular prices to be established.”179 By contrast, in NIRA, a Depression Era statute, Congress “prescribed no method of attaining that end save by the establishment of codes of fair competition, the nature of whose permissible provisions was left undefined” and “provided no standards to which those codes were to conform.”180 The Court observed that the “essentials of the legislative function are the determination of the legislative policy and its formulation and promulgation as a defined and binding rule of conduct.”181 The Constitution “does not demand the impossible or the impracticable.”182 A review of the Court’s treatment of challenges to federal and state statutes on the grounds that they had impermissibly delegated legislative power to nonlegislative actors does not provide much basis for thinking that there was ever a seriously confining nondelegation doctrine as part of the effective constitutional order. The New Deal cases of 1935 do not represent a routine effort by the Court to defend traditional separation-of-powers principles, and the constitutional revolution of 1937 did little to delineate a meaningful constitutional rule regarding the delegation of lawmaking authority. The Court had long recognized that legislative power as such could not be delegated, and the Court continued to maintain that idea after 1937. But the Justices simultaneously had long insisted that this principle had to be understood in a way that was practical and did not clog the machinery of modern government. The Court repeatedly emphasized that the nondelegation principle was no obstacle to the rise of the administrative state. The Court simply noted there had to be some limit as to how much authority Congress could delegate. Although the Justices tended to shy away from drawing a definite line around the power to delegate, the line they suggested was constantly in

177 Id. 178 Yakus v. United States, 321 U.S. 414, 423-24 (1944). 179 Id. 180 Id. at 424. 181 Id. 182 Id. 2017] The Myth of the Nondelegation Doctrine 405 retreat as Congress delegated more authority to other government officials and the Court sanctioned those moves. At the extreme, Congress could not simply pass off a policy domain to the executive branch with no indication of what the executive was to do in that authority. From that perspective, the New Deal cases look less like business as usual under the classical Constitution than an extraordinary instance of congressional abdication of exactly the sort that the Court had long cautioned against. When the Court once again began to approve New Deal delegations of regulatory authority to administrative agencies and executive officials, its actions suggest less that a part of the Constitution had been thrown into exile than the return to normalcy. With a modicum of specificity, Congress could provide the “adequate” policy guidance that the Court had long required for a statute to pass constitutional muster.183

C. The Doctrine in State and Lower Federal Courts Over the course of its history, the Supreme Court has addressed relatively few nondelegation challenges to statutes. The Court’s own explanation of the source and logic of the nondelegation principle has been thin, and the doctrine the Court has articulated to determine whether the delegation of policymaking authority is constitutionally permissible has been vague. While some other features of federal constitutional law that came under challenge in the New Deal period were robust and the subject of frequent deliberation and detailed analysis on the part of the Justices,184 the nondelegation doctrine largely languished in the shadows. Even so, the nondelegation principle was a familiar part of the constitutional landscape of the nineteenth century. The venerable Thomas Cooley stated plainly, One of the settled maxims in constitutional law is, that the power conferred upon the legislature to make laws cannot be delegated by that department to any other body or authority. Where the sovereign power of the State has located the authority, there it must remain; and by the constitutional agency alone the laws must be made until the constitution itself is changed.185 But Cooley was as likely to turn to the supreme courts of the states as to the Supreme Court of the United States when identifying the settled maxims

183 See Sunshine Anthracite Coal Co. v. Adkins, 310 U.S. 381, 398 (1940) (upholding the delegation of fixing reasonable prices for coal when the provided standard exceeded the previously upheld standard of “just and reasonable”). 184 See generally, e.g., Charles L.B. Lowndes, The Tax Decisions of the Supreme Court, 1938 Term, 88 U. PA. L. REV. 1 (1939) (discussing the doctrinal innovation regarding federal tax policy in the New Deal era and evaluating the volume of recent constitutional challenges to federal taxing power). 185 THOMAS M. COOLEY, A TREATISE ON THE CONSTITUTIONAL LIMITATIONS WHICH REST UPON THE LEGISLATIVE POWER OF THE STATES OF THE AMERICAN UNION 116-17 (Da Capo Press 1972) (1868). 406 University of Pennsylvania Law Review [Vol. 165: 379 of American constitutional law. The state supreme courts and lower federal courts had numerous opportunities to hear litigants argue that statutes had impermissibly delegated the legislative power to other government officials and to develop a judicial understanding of what the prohibition on the excessive delegation of lawmaking authority actually meant in practice. In this Section, we consider how state courts and lower federal courts elaborated the nondelegation doctrine from the Founding Era through the New Deal. These courts were more active than the U.S. Supreme Court, and often more articulate and elaborate in explaining the logic of the Constitution’s nondelegation principle. But these courts likewise refrained from imposing sharp limits on legislative discretion to shift important swaths of policymaking to other government officials, and they too erected few practical barriers to the rise of the modern administrative state and the expansive executive role in initiating and designing regulatory policy. In 1799, the Supreme Court of Pennsylvania heard America’s first nondelegation challenge.186 The statute at issue was passed by the Pennsylvania state legislature and empowered the city of Philadelphia to proscribe the construction of wooden homes as it “may judge proper.”187 In accordance with the statute, the city government enacted an ordinance banning the construction of wooden homes in a section of Philadelphia, and the defendant Philip Duquet was indicted for violating this ordinance.188 Duquet admitted to building a wooden home but argued that the city lacked the authority to make such a law.189 That power, he maintained, was the exclusive prerogative of the state assembly and could not be delegated to the city.190 In making his case, Duquet first noted that the Pennsylvania Constitution vested all legislative power in the general assembly.191 Next, he appealed to the common law maxim delegata potestas non potest delegari to persuade the court that this ordinance violated a fundamental constitutional principle—namely that power delegated by the people may not be redelegated to another institution.192 The people had entrusted the state assembly alone with the power to legislate, and any subsequent delegation by the assembly would be an abdication of its responsibility.193 Despite a lengthy argument by Duquet,

186 See Respublica v. Duquet, 2 Yeates 493 (Pa. 1799). 187 Id. at 492. 188 Id. at 492-94. 189 Id. 190 Id. at 493-94. 191 Id. at 494 (“[T]he first section of the first article [of the Pennsylvania Constitution] declares, that the legislative power shall be vested in a general assembly, which shall consist of a senate and house of representatives.”). 192 Id. at 494-96. 193 Id. 2017] The Myth of the Nondelegation Doctrine 407 the Pennsylvania Supreme Court disposed of the nondelegation question simply by stating, “We however see no such [constitutional] violation in the present case, and therefore give judgment for the commonwealth.”194 With that, the first nondelegation challenge was stopped in its tracks. It would be nearly three decades before any state court heard another case regarding the constitutionality of legislative delegation. When that time came, however, courts began to give more credence to nondelegation challenges. Given that the U.S. Supreme Court had taken up the issue of nondelegation in three separate cases following the Pennsylvania decision,195 it is not surprising that judges began to feel compelled to acknowledge the doctrine as an important principle and to provide arguments justifying their rulings. Unlike in Duquet, it was no longer acceptable to simply declare the statute constitutional without further explanation. The next case to reach the state courts, In re Adams,196 illustrates the beginnings of this shift. In Adams, decided in 1826, the Supreme Judicial Court of Massachusetts had to determine whether the state legislature could delegate to the governor and his council the power “to organize and arrange the militia of this commonwealth, conformably to the laws of the United States, and to make such alterations therein, as, from time to time, may be deemed necessary.”197 The court held that the delegation had to be constitutional because a contrary ruling “would lead to great difficulties and embarrassments” that would prevent the government from functioning effectively.198 In reaching this conclusion, the justices highlighted the fact that prior to delegating its authority to the governor, the state legislature had been delegated the authority to organize the militia by a higher power—the United States Congress.199 Because the Constitution requires Congress—not the states—“[t]o provide for organizing, arming and disciplining, the Militia,”200 if the state legislature’s delegation to the governor were unconstitutional, then Congress’s delegation to the state legislature would likewise be unconstitutional.201 This, the court held, would be an absurd result.202

194 Id. at 501. 195 See supra text accompanying notes 80–108. 196 In re Adams, 21 Mass. (4 Pick.) 25 (1826). 197 Id. at 28 (internal quotations omitted). 198 Id. at 29. 199 Id. at 28. 200 U.S. CONST. art. I, § 8, cl. 16. 201 See In re Adams, 21 Mass. (4 Pick.) at 29 (“[I]f the law of the state purporting to authorize the commander in chief, with advice of council, to organize and arrange the militia be void, being an improper delegation of authority, the act of congress providing that the militia of the respective states shall be arranged as the legislature of each state shall direct, must be considered void for the same reason.”). 202 See id. (“The consequence would be, that the most inconsiderable alteration in the organization of the militia could not be made without an express act of congress specially authorizing such alteration.”). 408 University of Pennsylvania Law Review [Vol. 165: 379

Notably, the justices did leave open the possibility that there could be unconstitutional delegations of power. In the court’s view, the salient factor was whether the original holder of authority ceded its “controlling power” to another body.203 In this instance, neither Congress nor the state legislature relinquished control. For that reason, the statute was constitutional. As courts began to work out more precise contours of the nondelegation doctrine, they frequently looked to constitutional and jurisprudential maxims to help define the permissible limits of delegation. The aforementioned maxim delegata potestas non potest delegari was one of the most influential.204 Although some courts saw this as a near absolute prohibition on delegation,205 most took a more nuanced view.206 As the Supreme Court of Michigan wrote in 1854, “it is in the very nature of legislative power, that it may, to some extent at least, be delegated, and that the maxim, delegata potestas, non potest delegari, has no application” in some cases.207 Likewise, the Missouri Supreme Court stated that “[a]lthough it be true, as a general proposition, that the legislature can not delegate their legislative power, but must exercise it themselves under their appropriate responsibilities,” there are many longstanding exceptions to this rule.208

203 See id. (observing that “[a]ll that was intended by the clause referred to, in the constitution of the United States, was, to give to congress a controlling power in organizing the militia”). 204 See, e.g., Franklin Bridge Co. v. Wood, 14 Ga. 80, 83 (1853) (“[T]he Legislature is but the agent of their constituents; and . . . they cannot transfer authority delegated to them to any other body . . . unless expressly empowered by the Constitution to do so. That to do this, would be to violate one of the fundamental maxims of jurisprudence, as well as of political science, namely: delegata potestas, non potest delegari.”); People ex rel. Caldwell v. Reynolds, 10 Ill. (5 Gilm.) 1, 11 (1848) (“To the general assembly have the people delegated the legislative powers of the government, only limited and controlled by the Federal and State constitutions, and it is insisted that these powers cannot be delegated to any body of men or any portion of the people, upon the principle that delegated powers can not be delegated. This maxim is true . . . .”); Thorne v. Cramer, 15 Barb. 112, 116 (N.Y. Gen. Term 1851) (“[A legislator] cannot delegate to others the trust which has been expressly confided to him, by reason of his supposed knowledge and sound judgment. Delegata potestas, non potest delegati, is a settled maxim of the common law, in full force at the present day; and never more applicable than to the case of a legislator . . . .”); Parker v. Commonwealth, 6 Pa. 507, 515 (1847) (“Among the primal axioms of jurisprudence, political and municipal, is to be found the principle that an agent, unless expressly empowered, cannot transfer his delegated authority to another, more especially when it rests in a confidence, partaking the nature of a trust, and requiring for its due discharge, understanding, knowledge, and rectitude. The maxim is, delegata potestas non potest delegari.”). 205 See, e.g., Thorne, 15 Barb. at 116 (citing the maxim in defense of its holding that the state legislature may not delegate to the voters the power to determine whether free schools shall be established in the state); Parker, 6 Pa. at 515-16 (relying on the maxim to conclude that it is unconstitutional for the legislature to delegate to the voters the power to determine whether the sale of liquor will be legal within certain counties). 206 See, e.g., Dubuque Cty. v. Dubuque & Pac. R.R. Co., 4 Greene 1, 2-3 (Iowa 1853) (discussing a variant of the maxim “Deleqare non delegatum est,” but ultimately deciding that it does not preclude the legislature from delegating to the voters the power to determine whether the county shall aid in the construction of a railroad). 207 People v. Collins, 3 Mich. 343, 368 (1854). 208 Wells v. City of Weston, 22 Mo. 384, 389 (1856). 2017] The Myth of the Nondelegation Doctrine 409

This idea that historical practice limited the scope of the nondelegation doctrine was very powerful.209 Even when the text of the state constitution seemed to enshrine the nondelegation doctrine by vesting all legislative power in the state assembly, many courts appealed to historical practice to justify allowing legislative delegations.210 Ultimately, when rejecting nondelegation challenges, courts generally took a pragmatic view of the situation and upheld those delegations that they deemed necessary for the government to accomplish its goals.211 Although such exceptions to the nondelegation doctrine were prevalent, they did not completely swallow the rule. The constitutional principle of separation of powers served as a counterweight against giving legislatures too much latitude.212 Courts knew that if the branches of government could delegate their powers however they saw fit, then the carefully constructed system of checks and balances would disintegrate, and the people would quickly find their liberties curtailed. An 1843 case decided by the Pennsylvania Supreme Court illustrates this concern: It is on the preservation of the lines which separate the cardinal branches of the government, that the liberties of the citizen depend; for a consolidated sovereignty, in whatever form, is a despotism in so far as it subjects the governed . . . and a government becomes consolidated in proportion as its

209 See Tilley v. Savannah, 5 F. 641, 657 (C.C.S.D. Ga. 1881) (holding that the delegation of certain “powers is not to be considered as trenching upon the maxim that legislative power is not to be delegated, since that maxim is to be understood in the light of the immemorial practice of this country and England, which has always recognized the propriety of vesting in municipal corporations certain powers of local regulation”); Wells, 22 Mo. at 389 (observing that it is so “firmly established, and daily practiced by our American governments when our constitution was adopted” that state legislatures must have the authority to delegate legislative power to municipal corporations); Thompson v. Floyd, 47 N.C. (2 Jones) 313, 315 (1855) (stating that the practice of the legislature delegating portions of its legislative functions “has been too long settled and acquiesced in by every department of the government and by the people, to be now disputed or even discussed”). 210 See, e.g., Collins, 3 Mich. at 349 (holding that despite the constitutional provision stating that “[t]he legislative power is vested in a senate and house of representatives,” the legislature may delegate to the people the power to decide whether an act prohibiting the sale of alcohol should become law). But see W. Union Tel. Co. v. Poe, 61 F. 449, 467 (C.C.S.D. Ohio 1894) (invalidating a statute that delegated taxing power to a board of appraisers in part because “Article 2, § 1, of the constitution provides that the legislative power of this state shall be vested in a general assembly, which shall consist of a senate and house of representatives”). 211 See Thompson, 47 N.C. (2 Jones) at 316 (“The truth is, that in the management of all the various and minute details, which a highly civilized and refined society requires, the General Assembly must have, and are universally conceded to have, the power to act by means of agents . . . . Without such power the Legislature would be an unwieldy body, incapable of accomplishing one-half of the great purposes for which it was created.”). 212 See, e.g., State v. Field, 17 Mo. 529, 532 (1853) (invalidating a statute that contained a provision permitting the courts to refuse to enforce a statute because such a legislative delegation violated the constitutional requirement that “the powers of government shall be divided into three departments, each of which shall be confided to a separate magistracy”). 410 University of Pennsylvania Law Review [Vol. 165: 379

legislative branch abandons its own functions, or usurps those which have been vested elsewhere.213 The people had entrusted specific powers to the legislatures for carefully considered reasons. When upholding the constitutionality of legislative delegations, courts needed to ensure that such delegations would not destabilize the delicately balanced separation of powers. In re Adams posited a solution to this problem.214 Specifically, the separation of powers principle does not mandate a complete prohibition of all delegations. Instead, it simply requires the original branch of government to retain “controlling power” over the given domain.215 The legislature is free to delegate its powers so long as the delegation provides clear guidelines.216 If the delegation provides another body with arbitrary discretion, then it fails this test and must be declared unconstitutional.217 In upholding a delegation that permitted the voters to set the location of the county seat, the Supreme Court of California gave one of the earliest articulations of this idea: By the Constitution the Legislature is required to provide for many objects which cannot be effected by the directaction of the Legislature, and while the maxim delegatus non potest delegare, is undoubtedly true, the extent of its application to legislative bodies must depend upon the nature and design of the legislation and the means necessary to accomplish the design, as well as a knowledge of the powers of the Legislature and the acts which may be done in the exercise of those powers.218 In other words, when assessing the constitutionality of a delegation, courts should look to the purpose of the legislation and the method by which the legislature has directed its agent to accomplish that purpose. If the legislature’s desired outcome were reasonably specified and the method of reaching that outcome seemed appropriate, these factors would weigh heavily in favor of finding the delegation constitutional. Despite being expounded in 1857, this standard is remarkably similar to the “intelligible principle” standard

213 In re Borough of W. Phila., 5 Watts & Serg. 281, 283 (Pa. 1843). 214 See supra notes 196–203 and accompanying text. 215 See In re Adams, 21 Mass. (4 Pick.) 25, 29 (1826) (noting that the Framers of the Constitution intended to give Congress a controlling power in organizing the militia either by organizing the militia themselves or directing others to organize it). 216 See Mistretta v. United States, 488 U.S. 361, 372-73 (1989) (holding that Congress could delegate power to a judicial branch commission charged with promulgating binding sentencing guidelines). 217 See Can. N. Ry. v. Int’l Bridge Co., 7 F. 653, 656 (N.D.N.Y. 1880) (upholding a statute delegating to courts the power to ascertain whether railroad companies have equal privileges in using a certain bridge because the statute intended for “this decision [to] proceed upon settled principles of law and equity, and not upon arbitrary discretion,” thereby suggesting that a legislature that granted arbitrary power in the course of delegation would act unconstitutionally). 218 Upham v. Supervisors of Sutter Cty., 8 Cal. 378, 382-83 (1857). 2017] The Myth of the Nondelegation Doctrine 411 adopted by the Supreme Court more than seventy years later in J.W. Hampton.219 Not only did the two have similar criteria—they also yielded similar outcomes. Much like the Supreme Court, the state and lower federal courts rarely met a principle that was not “intelligible.”220 But when they did, it was for the same reason given by the Supreme Court: the legislature had conferred too much discretion upon the delegatee.221 That said, determining precisely what amounts to too much discretion proved to be a hard task. In their attempts to resolve this issue, courts turned to several more tractable questions to measure whether this line had been crossed. First, was the group upon whom the legislature conferred power qualified to carry out the delegated task? And second, in delegating the power, was the legislature abdicating its responsibility to the public or shielding itself from electoral accountability? With respect to the first factor, courts were likely to find the statute unconstitutional when it delegated power to a group that was not fit to wield the power. Generally, this test was used to invalidate delegations to the electorate. Early on, many courts simply did not think voters were capable of wielding meaningful legislative authority.222 The Supreme Court of Delaware stated, [People are] incompetent . . . to exercise with discernment and discretion, collectively, or by means of the ballot-box, the power of legislation; because, under such circumstances, passion and prejudice incapacitate them for deliberation; and the tricks of demagogues, excited feelings, party animosities,

219 See supra text accompanying notes 151–60. 220 See supra text accompanying notes 186–211. 221 See supra text accompanying notes 212–18. 222 See Thorne v. Cramer, 15 Barb. 112, 117 (N.Y. Gen. Term 1851) (“The doctrine that no harm can result from allowing the people to exercise, directly, the law-making power, is more plausible than sound. . . . It is hardly necessary to say, that many voters are not in all respects qualified to become governors or legislators. They may have discretion enough to select suitable men for those offices; but if they were put directly to the business of framing laws themselves, they would be quite out of their element. Can we not then foresee dangers to arise from a delegation of the legislative franchise, even to the people themselves?”); Parker v. Commonwealth, 6 Pa. 507, 520 (1847) (“[I]f the two houses can divest themselves of their officeof law-makers, and devolve it upon the body of the people, what security have we against the passage of laws, perhaps well meant, but liable to be glaringly wrong, because inconsiderately adopted? [A]nd what check is left us upon hasty and ill-advised zeal, open to be influenced and misguided by interested, cunning, or blind fanaticism? If the practice be sanctioned, there may follow a train of experiments which, unarrested at some point of their progress, must end in the final overthrow of the constitution. Every case of doubtful propriety will be referred to the result of a ballot; and acts of Assembly, subject to the popular vote, will be yielded to unthinking clamor or partisan importunity . . . .”). But see State v. Parker, 26 Vt. 357, 364 (1854) (“Does any one seriously doubt the perfect propriety of the legislature, upon questions of general policy, affecting equally the whole state, acting upon the known will of the state, where that is known? We suppose not.”). 412 University of Pennsylvania Law Review [Vol. 165: 379

and the corrupting influences always brought to bear upon popular elections, would banish reason, reflection, and judgment.223 Although the exact limits of the competence of the electorate—or any other delegatee—can be disputed, the underlying principle is sound: to be the object of delegation, one must be able to exercise the power in a rational manner. If the delegatee is incapable of satisfying this condition, then the legislature has granted too much discretion. The second factor many courts considered was whether the legislature had abdicated its responsibility to the public. In particular, did the delegation permit so much discretion that the legislature could reasonably disclaim ownership of the law and thereby shield itself from electoral accountability? If it did, that would weaken the constitutionally mandated connection between legislative action and electoral accountability that is at the very core of the American political system. Ensuring that the appropriate branch of government was answerable for its actions was so important that many judges feared “shifting responsibility [would] introduce innovations upon our system, which would result in the overthrow and ultimate destruction of our political fabric.”224 Courts were certain that, if left unchecked, “faithless legislators anxious to escape the responsibility of their position” would abuse delegation to ensure their political survival.225 Such action, the courts knew, would be a severe breach of the trust that the people had placed in their elected representatives.226 In advancing the accountability argument, courts also analogized to other branches of government by observing that neither the executive nor the judiciary could delegate away their powers. In fact, to even contemplate such a possibility would be “absurd.”227 As one New York state court wrote,

223 Rice v. Foster, 4 Del. (4 Harr.) 479, 489-90 (1847). 224 Franklin Bridge Co. v. Wood, 14 Ga. 80, 83 (1853). 225 Parker, 6 Pa. at 520. 226 See, e.g., Moore v. Allen, 30 Ky. (7 J.J. Marsh.) 651, 652 (1832) (“The power confided to members of congress is a personal trust, which can not be transferred by them. When called on to account to their constituents for their conduct, it would be at war with our whole system to excuse [legislators] upon the ground that they had delegated their powers to [another body].”). 227 See Rice, 4 Del. (4 Harr.) at 489 (“If the legislative functions can be transferred or delegated to the people, so can the executive or judicial power. The absurd spectacle of a governor referring it to a popular vote, whether a criminal, convicted of a capital offence, should be pardoned or executed, would be the subject of universal ridicule: and were a court of justice, instead of deciding a case themselves, to direct the prothonotary to enter judgment for the plaintiff, or defendant, according to the popular vote of a county, the community would be disgusted with the folly, injustice, and iniquity of the proceeding.”). 2017] The Myth of the Nondelegation Doctrine 413

[T]he same reasoning that would permit a legislator to transfer his power to his constituents, would authorize any other elective officer to do the same. Thus the governor when applied to, to pardon a criminal, might, being unwilling to take the responsibility of deciding upon the application himself, call an election and submit it to the people. The courts, whenever a case of peculiar difficulty came before them, might call together their constituents to ascertain how the popular feeling stood. . . . Every person must perceive how preposterous such a proceeding would be! And how deservedly contemptible every officer and court who should resort to such means of evading the just responsibilities of his office, would be held.228 In one egregious instance of delegation, the Missouri state legislature passed a law that gave the county courts complete discretion to suspend the operation of that law.229 The statute contained no limiting principle. If at any time, the court determined “that the provisions of the act should not be enforced,” it could simply decline to enforce them.230 In striking down the delegation as unconstitutional, the Missouri Supreme Court emphasized that it is “the duty of the legislature to exercise the powers upon their own responsibility. . . . [T]he power, thus committed by the people into the hands of their constitutional representatives, is not to be delegated to others not trusted by the people.”231 In another case, the Tennessee legislature delegated to courts the power to set taxes “upon all polls and property subject to taxation by the laws of this State.”232 The legislature did not specify rates of taxation or even whether certain types of property should be taxed more or less heavily than any other. The only requirement was that the courts assess taxes sufficient “to meet the current expenses of their county for the ensuing year.”233 In reviewing the constitutionality of the delegation, Tennessee’s highest court asked, [W]hat limit to exactions is imposed by the act . . . ? We answer, none. [The courts] may tax every acre in their respective counties to its full value, and if the tax is not paid, cause the land to be sold and bought in by the sheriff . . . if there be no other bidders.234

228 Thorne v. Cramer, 15 Barb. 112, 116-17 (N.Y. Gen. Term 1851). 229 State v. Field, 17 Mo. 529, 530 (1853). 230 Id. 231 Id. at 536. 232 Marr v. Enloe, 9 Tenn. (1 Yer.) 452, 453 (1830). 233 Id. 234 Id. at 454. 414 University of Pennsylvania Law Review [Vol. 165: 379

Under this system, tax rates are set solely by the caprice of judges.235 Such a delegation, the court ruled, was unconstitutional. These cases are not outliers. Rather, they are representative of the kinds of delegations that state and lower federal courts held to be unconstitutional during the nineteenth century. If anything about the nondelegation doctrine’s history is surprising, it is that legislatures once thought it appropriate to delegate such expansive powers, not that courts saw fit to strike them down. Since the beginning, the limits of the nondelegation doctrine have remained constant. To the extent that its influence appears diminished, it is because the set of cases is different. If this year, Congress were to delegate all taxing power to the judiciary or entrust judges to determine whether a law should be enforced, there is no doubt the courts would hold those statutes to be unconstitutional. Modern judges, however, are not confronted with such profoundly unconstitutional delegations. Instead, today’s disputes hew much closer to the constitutional line. In delineating the contours of the nondelegation doctrine, courts justified their decisions by appealing to three forms of support: (1) precedent, (2) maxims, and (3) constitutional text. Table 1 illustrates the percentage of federal and state cases that cited to each of these forms of support. As the data show, courts most frequently invoked precedent, followed by maxims, and finally constitutional text. The Table further breaks down the cases into those that found the delegation to be constitutional and those that found it to be unconstitutional. Unsurprisingly, when striking down a statute, both state and federal courts were more likely to cite a precedent, invoke a maxim, or quote constitutional text to support their ruling. A chi-square test also reveals that federal courts were significantly more likely than state courts to cite precedent when striking down a statute and that state courts were significantly more likely than federal courts to quote constitutional text regardless of whether the statute was ruled invalid.

235 Id. at 454-55 (“Until county courts by its order (clearly amounting to a legislative act) imposes the tax, the people have no knowledge what they have to pay; nor have they any knowledge afforded them, even by the order fixing the tax . . . .”). 2017] The Myth of the Nondelegation Doctrine 415

Table 1: Percentage of Constitutional and Unconstitutional Delegations for Which the Federal and State Courts Cited Support

Precedent Maxim Constitution

Con. Uncon. Con. Uncon. Con. Uncon. Federal 56% 88% 19% 44% 6% 27% State 54% 68% 19% 32% 20% 43%

χ2 statistic 0.44 6.96** 0.0007 2.21 36.19*** 4.19* * p < .05; ** p < .01; *** p < .001

Because we have already discussed the importance of maxims and precedent in nondelegation cases, we focus our attention here on the differing role constitutional text plays in state and federal cases. We highlight one distinction in particular: whereas the nondelegation doctrine is only implicit in the U.S. Constitution, it is explicit in most state constitutions. As Gary Lawson has observed, “there is nothing in the [U.S.] Constitution that specifically states, in precise terms, that no other actor may exercise legislative power or that Congress may not authorize other actors to exercise legislative power.”236 Accordingly, given the lack of a specific nondelegation clause, scholars have derived support for the doctrine from Article I, Section 1 of the U.S. Constitution. This provision reads, “All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives.”237 From here, federal courts have inferred the existence of the nondelegation doctrine,238 reading this Section to require that judges “guard jealously against any attempt of the legislative authority to delegate its power to others.”239

236 Lawson, supra note 4, at 335; see also Cynthia R. Farina, Statutory Interpretation and the Balance of Power in the Administrative State, 89 COLUM. L. REV. 452, 478 (1989) (stating that the nondelegation doctrine is not expressly grounded in the text of the Constitution); Cass R. Sunstein, Nondelegation Canons, 67 U. CHI. L. REV. 315, 322 (2000) (indicating that the text of the Constitution does not provide unambiguous support for the conventional nondelegation doctrine). 237 U.S. CONST. art. I, § 1. 238 See, e.g., Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 42-43 (1825) (“It will not be contended that Congress can delegate to the Courts, or to any other tribunals, powers which are strictly and exclusively legislative.”); see also Andreas v. Clark, 71 F.2d 908, 910 (9th Cir. 1934) (citing Wayman v. Southard for the proposition that Congress cannot delegate powers which are exclusively legislative); United States v. Griffin, 12 F. Supp. 135, 136 (S.D. Ga. 1935) (invoking Article I, Section 1 of the U.S. Constitution to strike down an act that “attempts unlawfully to delegate legislative authority to the Executive Department of the United States”). 239 United States v. Edwards, 14 F. Supp. 384, 393 (S.D. Cal. 1936) (citing Marshall Field & Co. v. Clark, 143 U.S. 649 (1892)). 416 University of Pennsylvania Law Review [Vol. 165: 379

Unlike the U.S. Constitution, most state constitutions make direct reference to the nondelegability of legislative powers and do so in a number of ways. First, they emphasize that the state government is a system of separation of powers and, as such, no branch of government may exercise any powers that are within the proper domain of another branch. Article II, Section 1 of the Texas constitution is a representative example. That provision states, “The powers of the Government of the State of Texas shall be divided into three distinct departments . . . and no person, or collection of persons, being of one of these departments, shall exercise any powers properly attached to either of the others.”240 Second, many state constitutions forbid the legislature from making the passage of any law contingent upon any event or outside authority. For example, Article I, Section 25 of the Indiana constitution has such a provision. It reads as follows: “No law shall be passed, the taking effect of which shall be made to depend upon any authority, except as provided in this Constitution.”241 Finally, more than a dozen states explicitly forbid the legislature from delegating any of its powers. The language in Article V, Section 35 of the Colorado constitution is representative: “The general assembly shall not delegate to any special commission, private corporation or association, any power to make, supervise or interfere with any municipal improvement, money, property or effects, whether held in trust or otherwise, or to levy taxes or perform any municipal function whatever.”242

240 TEX. CONST. art. II, § 1; see also ILL. CONST. art. II, § 1 (“The legislative, executive and judicial branches are separate. No branch shall exercise powers properly belonging to another.”); S.C. CONST. art. I, § 8 (“In the government of this State, the legislative, executive, and judicial powers of the government shall be forever separate and distinct from each other, and no person exercising the functions of one of said departments shall assume or discharge the duties of any other.”); VA. CONST. art. III, § 1 (“The legislative, executive, and judicial departments shall be separate and distinct, so that none exercise the powers properly belonging to the others, nor any person exercise the power of more than one of them at the same time . . . .”); WYO. CONST. art. II, § 1 (“The powers of the government of this state are divided into three distinct departments . . . and no person . . . charged with the exercise of powers properly belonging to one of these departments shall exercise any powers properly belonging to either of the others . . . .”). 241 IND. CONST. art. I, § 25; see also OR. CONST. art. 1, § 21 (“[N]or shall any law be passed, the taking effect of which shall be made to depend upon any authority, except as provided in this Constitution . . . .”). 242 COLO. CONST. art. V, § 35; see also CAL. CONST. art. XI, § 13 (1879) (“The legislature shall not delegate to any special commission, private corporation, company, association, or individual, any power to make, control, appropriate, supervise, or in any way interfere with, any county, city, town, or municipal improvement, money, property, or effects, whether held in trust or otherwise, or to levy taxes or assessments, or perform any municipal functions whatever.”). For an application of this provision, see Mesmer v. Bd. of Pub. Serv. Comm’rs of L.A., 138 P. 935, 935-36 (Cal. Dist. Ct. App. 1913), which applies this provision of the California Constitution of 1879 to limit the general lawmaking powers of the Legislature. See also PA. CONST. art. III, § 31 (“The General Assembly shall not delegate to any special commission, private corporation or association, any power to make, supervise or interfere with any municipal improvement, money, property or effects, whether held in trust or otherwise, or to levy taxes or perform any municipal function whatever.”); UTAH CONST. 2017] The Myth of the Nondelegation Doctrine 417

This explicit reference to the nondelegability of legislative powers is a far cry from the implicit references found in the U.S. Constitution. By including these provisions, states have made clear that the nondelegation doctrine is a valuable tool that courts should use to maintain the separation of powers between the three branches. Given the more extensive discussion of legislative delegations in state constitutions, it should come as no surprise that state courts almost never invoke the U.S. Constitution during their discussions of the nondelegation doctrine. The surprise is that even with these more robust textual supports for the nondelegation doctrine, state courts have still proved willing to defer to legislative assessments of when delegating power is practical and necessary.

III. JUDICIAL PRACTICE IN NONDELEGATION CASES There is little question that American courts have long recognized a basic constitutional principle that legislative powers cannot be delegated to other political actors. Having received the legislative power from the sovereign people, the elected representatives sitting in a legislature were expected to exercise it themselves. But that basic principle was immediately hemmed in by qualifications. Legislators were allowed to “commit something to the discretion of the other departments” and let the latter “fill up the details” of government policy.243 Legislators could reasonably turn to a “body of experts” when filling in the policy details themselves would be “impracticable.”244 Courts should not force legislatures into “great difficulties and embarrassments.”245 The principle of nondelegation had “to be understood in the light of the immemorial practice of this country,” which frequently tolerated the delegation of lawmaking power to other entities.246 In short, the nondelegation principle was never understood to impose substantial burdens on the legislative branch. Courts routinely expressed wariness as they approached the “delicate and difficult” task of identifying the boundary of the legislative power to bestow authority on their agents.247

art. VI, § 28 (“The Legislature shall not delegate to any special commission, private corporation or association, any power to make, supervise or interfere with any municipal improvement, money, property or effects, whether held in trust or otherwise, to levy taxes, to select a capitol site, or to perform any municipal functions.”); WYO. CONST. art. III, § 37 (“The legislature shall not delegate to any special commissioner, private corporation or association, any power to make, supervise or interfere with any municipal improvements, moneys, property or effects, whether held in trust or otherwise, to levy taxes, or to perform any municipal functions whatever.”). 243 Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 43, 46 (1825). 244 St. Louis Consol. Coal Co. v. Illinois, 185 U.S. 203, 211 (1902). 245 In re Adams, 21 Mass. (4 Pick.) 25, 29 (1826). 246 Tilley v. Savannah, 5 F. 641, 657 (C.C.S.D. Ga. 1881). 247 Wayman, 23 U.S. (10 Wheat.) at 46. 418 University of Pennsylvania Law Review [Vol. 165: 379

It is one thing to make a statement of constitutional principle; it is another to judicially enforce constitutional limits. The proto-realist Oliver Wendell Holmes famously characterized the formal law and the reports of judicial opinions as simply the materials from which “systematized prediction” of future judicial behavior can be made.248 Empirical legal scholars have long taken these “prophecies of what the courts will do in fact” as their central concern.249 The movement to study “judicial behavior” has been motivated by a conviction that judicial opinions and doctrines do not fully capture the reality of judicial practice, and that the “impressionistic focus on legal doctrine” is inadequate for understanding how political power is actually exercised.250 In this Part, we turn our attention to judicial practice in the state and federal courts from the Founding to the New Deal. Nostalgia for a constitution-in-exile is ultimately concerned less with abstract statements of principle than with the practical realities of constitutional limitation. Did judges, in fact, act to enforce a robust nondelegation constraint on the power and discretion of legislators? If so, under what circumstances? Is there evidence that the nineteenth-century version of the nondelegation doctrine had actual teeth? We begin by surveying the range of nondelegation cases in the state and federal courts between the Founding and the constitutional revolution of the New Deal period.251 Unless otherwise noted, the analysis in this Part draws on a dataset of 2506 cases decided in federal courts and state supreme courts between the Founding and 1940. These cases all involved challenges to legislative provisions on the grounds that the legislature inappropriately delegated lawmaking authority. A total of 421 cases resulted in the partial or total invalidation of a statutory provision during this period. Over eighty-five percent of the total number of cases were resolved in state supreme courts, and ninety percent of the cases striking down legislation took place in state courts. The constitutional law and practice of the nondelegation doctrine was thus largely written in the states. Figure 1 tracks the total set of cases addressing nondelegation constitutional challenges to legislative action across time and the number of cases in which

248 Oliver Wendell Holmes, Jr., The Path of the Law, 10 HARV. L. REV. 457, 458 (1897). 249 Id. at 461. 250 S. Sidney Ulmer, Judicial Review as Political Behavior: A Temporary Check on Congress, 4 ADMIN. SCI. Q. 426, 427-28 (1960). 251 In order to identify nondelegation cases, we conducted the following search on Westlaw for all state and federal cases between 1789 and 1940: “TO(‘delegation #of powers’) or (delegat! /2 legislative /1 (power! or authority)) or (delegat! /2 lawmaking /1 (power! or authority)).” We then proceeded to examine every search result to see if the case involved a nondelegation challenge. If it did not, we excluded the case from the dataset. If it did, we coded the case along a variety of dimensions which we discuss throughout this Part. 2017] The Myth of the Nondelegation Doctrine 419 the court struck down a legislative provision as unconstitutionally delegating lawmaking power.252 The figure shows the annual count of the number of cases.

Figure 1: Nondelegation Cases in State and Federal Courts, 1825–1940

120

100

80

60

40 Number of Cases

20

0 1825 1835 1845 1855 1865 1875 1885 1895 1905 1915 1925 1935 Year

All Nondelegation Cases Unconstitutional Delegations

As the figure makes evident, nondelegation cases were a regular feature of constitutional litigation prior to the New Deal, but there were substantial changes in their incidence over time. With two exceptions,253 the first cases to challenge legislation on the grounds that the legislature inappropriately delegated lawmaking authority did not emerge until the 1820s. Even then, however, such cases remained relatively rare until the antebellum period. Over the course of the second half of the nineteenth century, nondelegation cases made a regular appearance on judicial dockets, gradually increasing in the last years of the Gilded Age. Nondelegation cases surged at the opening of the twentieth century, plateauing at a new level that was several times the pace at which such cases were heard in the nineteenth century. In keeping with the traditional narrative of the battles of the New Deal, the number of

252 Because only two nondelegation challenges arose prior to 1825, as a practical matter, the graphs focuses on cases between 1825 and 1940. 253 See Cargo of the Brig Aurora v. United States, 11 U.S. (7 Cranch) 382 (1813); Respublica v. Duquet, 2 Yeates 493 (Pa. 1799). 420 University of Pennsylvania Law Review [Vol. 165: 379 nondelegation cases surged again in the 1930s. By the early twentieth century, nondelegation cases were a familiar feature of the American constitutional environment. Despite the growth of nondelegation as an area of litigation, the number of judicial invalidations hardly budged. As Figure 1 illustrates, the number of cases in which a court struck down a statutory provision in a case raising a nondelegation challenge remained on a nearly flat trajectory from the early nineteenth century through the early twentieth century. As litigation surged in the early decades of the twentieth century, a massive gap emerged between the number of cases bringing such challenges and the number of cases that were successful. The early 1930s do, however, stand out as an outlier, with a brief eruption of cases striking down legislation—an eruption that subsided as quickly as it arose. Although the number of cases resulting in an invalidation averaged less than one per year prior to 1880, that number remained in the low single digits until the 1930s. The number of nondelegation cases changed dramatically over the course of this period, as did their composition. Table 2 describes the delegatee of the lawmaking authority (or the object of delegation) at issue in these cases. It divides the data between two periods that largely map onto the change in the equilibrium number of cases identified in Figure 1: cases resolved before 1880 are compared to those resolved after 1880.254 The delegatee of legislation is divided into six categories: executive (primarily the chief executive), agency (including independent commissions and other bureaucratic bodies), judiciary, local government, voters, and others.255 Table 2 indicates some notable points of continuity in how legislatures have sought to delegate power to other political actors over time. The Chief Executive has always been one obvious possible object of delegation. The classic case of Field v. Clark, for example, involved a delegation to the President to determine whether the conditions for a higher tariff rate had been met.256 Somewhat surprisingly, given the prominence of cases involving the President in the U.S. Supreme Court, such delegations have always been relatively rare. The judiciary has been a more common recipient of delegated

254 The transformation described in Table 2 evolved gradually over the final years of the nineteenth century and little of substance would be changed by choosing a somewhat different point at which to divide the data. These developments are more clearly represented in a table, however, than a figure. 255 The catch-all “other” category includes such idiosyncratic objects as state legislatures, see, e.g., Moore v. Allen, 30 Ky. (7 J.J. Marsh.) 651, 652-53 (1832) (evaluating a federal statute that gave federal prisoners the same privileges as possessed by the local state prisoners); surrounding property owners, see, e.g., City of Chicago v. Stratton, 58 Ill. App. 539, 544-46 (1895) (evaluating a law authorizing neighboring residents to allow or prohibit the location of a livery stable); and private corporations, see, e.g., Smith Agric. Chem. Co. v. Calvert, 18 Ohio Dec. 583, 587-88 (Ohio Ct. Com. Pl. 1908) (evaluating a state statute that authorized the board of agriculture to exercise regulatory powers). 256 Field v. Clark, 143 U.S. 649, 692-96 (1892). 2017] The Myth of the Nondelegation Doctrine 421 power, as when the Georgia legislature in 1843 delegated to the local courts the task of creating corporations.257 But cases questioning such delegations are equally represented both early and late in this time period.

Table 2: Objects of Delegation

Local Executive Agency Judiciary Gov’t Voters Other Before 1880 6% 3% 15% 27% 47% 2% After 1880 5% 52% 13% 24% 8% 10%

χ2 *** *** *** statistic 0.04 122.63 3.01 1.02 297.43 13.96 *** p < .001

An extremely prominent body of nondelegation cases at the state level is virtually unknown at the federal level—delegations from state legislatures to local governmental units. Cases determining whether state legislatures can delegate a general police power to municipal corporations258 or may delegate the power to choose the siting of county buildings to county commissioners259 have been common at the state level, shaping the judicial understanding of the meaning of the nondelegation principle and its exceptions. Such cases have occupied a relatively steady place on state judicial dockets across the decades. Table 2 also directs our attention to two important points of discontinuity between the delegation in the early nineteenth century and delegation since the late nineteenth century. The first is less surprising from a modern perspective. Legislative delegations of rulemaking authority to executive agencies gave rise to a trivial number of cases prior to 1880. The use of independent regulatory commissions and other specialized, expert bureaucratic units was pioneered during the Gilded Age, however, and these new devices of governance and administration brought with them new constitutional disputes. One early example was the creation of the Railroad and Warehouse Commission in Illinois in 1871. Pursuant to the legislature’s authority under the state constitution to pass laws for the inspection of grain and the protection of

257 See Franklin Bridge Co. v. Wood, 14 Ga. 80, 84-85 (1853) (holding that conferring a mandatory duty on courts to give “legal form to these companies” did not unconstitutionally delegate a discretionary legislative power). 258 See, e.g., Sluder v. St. Louis Transit Co., 88 S.W. 648, 650-51 (Mo. 1905) (striking down a law that delegated power to private companies to set speed limits for street cars). 259 See, e.g., Simpson v. Bailey, 3 Or. 515, 518 (1869) (upholding an act of the legislature that authorized three commissioners selected by a county court to choose the site of county buildings). 422 University of Pennsylvania Law Review [Vol. 165: 379 producers, shippers, and receivers, the legislature empowered the three-person commission to set up a system of regulation and inspection of grain in the state.260 Even by 1878, the state supreme court thought it “now too late to question the power to create such agencies in the administration of the government, and invest them with such legislative power as shall be appropriate and necessary to effectuate the objects of their creation.”261 From the court’s perspective, there was little difference between the creation of a commission to exercise some of the police powers of the state legislature and the creation of other corporate bodies such as the “[c]ities, towns, villages, counties, townships, road districts and school districts” that were familiar features of the political landscape.262 The most controversial delegation to such railroad commissions in the years after the Civil War, however, was the authority to set carriage rates. But even here, courts were accommodating. The Minnesota Supreme Court, for example, concluded that all that was necessary was for the legislators to determine whether rates should be regulated.263 If the “sovereign state” were not to “find itself helplessly entangled in the meshes of its own constitution,” then it was necessarily the case that “the legislature may authorize others to do things which it might properly, but cannot conveniently or advantageously, do itself.”264 The legislative delegation of authority to an agency is the classic concern of modern disputes over the nondelegation doctrine and the central target of criticisms of the modern nondelegation doctrine.265 Both the absolute number and the relative share of nondelegation cases involving executive agencies steadily grew from the 1870s to the New Deal. The growth of such cases—and the underlying statutes empowering such institutions—was a primary engine behind the surge of nondelegation cases in the early twentieth century observed in Figure 1. The second discontinuity is less familiar since it marks a once common category of nondelegation cases that subsequently retreated into near invisibility. Moreover, these cases were endemic to state politics but largely absent from

260 People v. Harper, 91 Ill. 357, 364-65 (1878). 261 Id. at 366. 262 Id. 263 State ex rel. R.R. & Warehouse Comm’n v. Chi., M. & St. P. Ry. Co., 37 N.W. 782, 787 (Minn. 1888). 264 Id. 265 See, e.g., Lawson, supra note 6, at 1241 (arguing that “the demise of the nondelegation doctrine, which allows the national government’s now-general legislative powers to be exercised by administrative agencies, has encountered no serious real-world legal or political challenges, and none are on the horizon”); see also SCHOENBROD, supra note 66, at 155-64 (arguing that the New Deal did not amend the Constitution to permit congressional delegation to agencies and that such delegation is unconstitutional). See generally PHILIP HAMBURGER, IS ADMINISTRATIVE LAW UNLAWFUL? (2014) (framing the modern debate on the nondelegation doctrine and contending that modern American administrative law is unlawful). 2017] The Myth of the Nondelegation Doctrine 423 federal politics and the development of federal constitutional law. This class of cases involved delegations from legislators to voters. Although such cases were few in number, they occupied a large share of the nondelegation cases heard in the early decades of the nineteenth century. In the mid-nineteenth century, state legislators often turned to the innovative device of the referendum to empower local popular majorities to make such controversial policy decisions as whether to impose a school tax266 or prohibit the sale of alcohol.267 Judges were, in turn, called upon to decide whether the state legislature could pass the buck in this fashion. Despite this variety, all these cases were understood to implicate the same basic principle of American constitutional law—the extent to which legislatures could delegate power to other entities. The core concern was a consistent one of trying to understand what decisionmaking authority had to be retained by the legislature itself and what could be parceled out to others, and judges in all of these cases had to tread a similar path in trying to interpret the meaning and significance of the constitutional grant of legislative power to the legislature. The later judicial analysis of—and accommodation to—the legislative use of executive agencies built on the back of earlier analysis of—and accommodation to—the legislative use of voters and local governmental entities. Statutes delegating authority to nonlegislative actors varied not only by the recipient of the delegated authority but also by the type of authority being received. Table 3 places these nondelegation cases into four categories depending on the type of law at issue. It likewise organizes the data into two time periods, before and after 1880. Cases are distinguished by whether the delegated power in question primarily involves regulation, taxation, spending, or something else (i.e., other).

Table 3: Subject Matter of Delegation

Regulation Taxation Spending Other

Before 1880 28% 34% 6% 33%

After 1880 54% 14% 8% 25%

χ2 statistic 52.07*** 54.30*** 0.8491 7.392** ** p < .01; *** p < .001

266 See, e.g., Steward v. Jefferson, 3 Del. (3 Harr.) 335, 336-37 (1841) (finding a state statute constitutional that permitted a tax to be imposed upon residents of a school district following a vote by those residents). 267 See, e.g., Parker v. Commonwealth, 6 Pa. 507, 514-15 (1847) (holding that state legislatures could not constitutionally delegate to voters the power to determine whether the sale of alcohol will be legal). 424 University of Pennsylvania Law Review [Vol. 165: 379

Again, there are important points of continuity and discontinuity. Delegations of spending authority occupy a small but persistent share of the nondelegation cases across this period.268 The catch-all “other” category is quite large here, often capturing the delegation of authority to manipulate political structures themselves—such as the congressional authorization of the state executive council to participate in the organization of the state militia269 or legislative authorization of the local voters to decide whether a new township should be preserved or dissolved.270 And again, two types of cases exchange places in their relative prominence on the docket. Cases involving the delegation of taxation authority were once the most notable subject matter involved in nondelegation cases. Who could impose taxes and financial obligations on the taxpayers (and for what purposes) were hot-button issues in the Jacksonian period, and courts were frequently asked to determine questions such as whether the legislature could authorize county courts to impose taxes to pay for road construction271 or whether the implementation of a law raising taxes for free public schools could be made contingent on the approval of the voters.272 Such cases, however, were in steady decline in the years after the Civil War, and a different category of cases more than filled their place. As Table 3 shows, by the beginning of the twentieth century, over half of the nondelegation cases involved legislative delegation of regulatory power. Of course, the rise of regulatory cases went hand-in-hand with the rise of cases involving executive commissions and agencies. In some instances, legislatures turned to more traditional executive officials, such as the U.S. Secretary of War, who was authorized to make any necessary rules and regulations to protect improvements on the Mississippi River.273 Riding circuit, Justice Lamar thought the crucial fact was not whether Congress had itself established the substance of the rules and regulations that would govern river traffic but that Congress “denounces the violation of it as a crime, and

268 See, e.g., City of Des Moines v. Hillis, 8 N.W. 638, 641 (Iowa 1881) (holding that the state legislature may authorize a municipal government to set a salary for a police judge). 269 See In re Adams, 21 Mass. (4 Pick.) 25, 29 (1826) (upholding a 1792 act of Congress allowing state governments to arrange militias as the legislature deemed appropriate). 270 See Commonwealth v. Judges of Quarter Sessions, 8 Pa. 391, 395-96 (1848) (finding such delegation constitutional). 271 See Justices of Clark Cty. Court v. Paris, Winchester & Ky. River Tpk. Co., 50 Ky. (11 B. Mon.) 143, 151-52 (1850) (finding such delegation of taxing authority constitutional). 272 See Johnson v. Rich, 9 Barb. 680, 681-82 (N.Y. Gen. Term 1851) (finding the submission of a tax to referendum to be permissible), overruled by Barto v. Himrod, 8 N.Y. 483, 489-91 (1853). 273 See United States v. Breen, 40 F. 402, 402 (C.C.E.D. La. 1889) (addressing whether such congressional delegation to the Secretary of War was constitutional). 2017] The Myth of the Nondelegation Doctrine 425 prescribes the penalty.”274 “[C]riminality” could only “result directly and exclusively from the legislation of congress,” but the establishment of rules that might generate such criminal violations could be left to others.275 Legislatures, however, were increasingly turning to newly created institutions, such as professional licensing boards. The Oregon Supreme Court, for example, found no difficulty with the eationcr of a state licensing board for barbers.276 The legislature had done its job in defining what constitutes a barber and specifying that those practicing the trade must receive a license from a state board. It could appropriately be left to the appointed members of the board of examiners to set all standards for determining who might qualify for a license.277 Similarly, the New Hampshire Supreme Court found that the legislature could leave to the state board of health the power to make “all necessary rules and regulations” which might advance the goal of preventing unhealthy conditions in food production, storage, and trade.278 “[G]eneral statutory authority” was sufficient; everything else could be “referred to some designated ministerial officer or body.”279 Since Reconstruction, nondelegation cases have largely revolved around the structure of the modern regulatory state, but it is important to recognize that such cases emerged out of—and continued alongside—a mix of other types of cases. The problem of nondelegation was not a unique result of the creation of a modern administrative bureaucracy but has been a persistent adjunct of the evolving challenges of governing with legislatures constituted by fundamental law. Federal constitutional law has been particularly concerned with delegation of regulatory power to executive agencies; however, within the development of American constitutional law, the federal courts have been relative latecomers and less active participants. Over the course of the nineteenth century, state courts were working through the principles of nondelegation in a more complex array of cases while federal courts were largely sitting on the sidelines. Recognizing this more complex legal environment within which the nondelegation doctrine developed and was applied prior to the New Deal places modern arguments about the nondelegation doctrine in a better context. The presumably robust nondelegation doctrine of the nineteenth century is often held as a significant obstacle to the kinds of progressive reforms advanced in the early twentieth century. The actual history of the nondelegation doctrine gives little support for that expectation.

274 Id. at 404. 275 Id. 276 State v. Briggs, 77 P. 750, 750-51 (Or. 1904). 277 Id. 278 State v. Normand, 85 A. 899, 900 (N.H. 1913). 279 Id. at 902. 426 University of Pennsylvania Law Review [Vol. 165: 379

As Figure 1 showed, the actual invalidation rate of litigated cases raising nondelegation challenges to legislation was generally low. Moreover, the odds of success became even longer as the modern regulatory state was being constructed and legislators experimented with such innovative tools as specialized regulatory commissions. Table 4 elaborates on those developments, indicating the rate at which the legislature lost these nondelegation challenges. Table 4 distinguishes between cases resolved in state courts and those resolved in federal courts. As indicated earlier, the state courts were by far the more common venue for hearing nondelegation cases.280 The invalidation rate in Table 4 reflects that reality, as only a handful of nondelegation cases were resolved by federal judges prior to 1880—and all were decided in the government’s favor. After 1880, the federal courts heard such cases more routinely as the administrative state was built and also began to rule against Congress. The state courts were more active in both the early and late nineteenth century in hearing such cases and in ruling against legislative efforts to delegate power.

Table 4: Invalidation Rate in State and Federal Cases

State Cases Federal Cases

Total 18% 12%

Before 1880 24% 0%

After 1880 17% 12%

While the invalidation rate in the state courts indicates a robust nineteenth century tradition of evaluating legislative action against accepted constitutional prohibitions on delegations of lawmaking power, it also suggests that the courts increasingly accommodated legislative innovations. As courts heard more nondelegation challenges in the late nineteenth and early twentieth centuries, federal judges finally found some cases in which the legislature had gone too far—but even then the invalidation rate was fairly low. The more active state courts generated more stable patterns of behavior that were less susceptible to the idiosyncrasies of individual cases, but on average these courts were less likely to strike down statutory provisions later in the period. Neither the state nor the federal courts were much of an obstacle to the delegation of legislative power to nonlegislative actors.

280 See supra notes 265–67 and accompanying text. 2017] The Myth of the Nondelegation Doctrine 427

Table 5: Invalidation Rate in All Cases by Object of Delegation

Local Executive Agency Judiciary Gov’t Voters Other

Total 20% 13% 21% 15% 17% 29% Before 1880 18% 20% 32% 24% 21% 17% After 1880 20% 13% 20% 14% 15% 29%

The declining willingness of even state courts to limit legislative delegations in part reflects the shifting composition of the nondelegation cases heard by the courts over time. As Table 2 indicates, delegations to executive agencies increased from a trivial proportion of early nineteenth-century nondelegation cases to a majority of the post-1880 cases. Conversely, cases delegating power to voters followed the opposite trajectory. Table 5 shows the invalidation rates across these different types of nondelegation cases. Statutes delegating power did not all have an equal likelihood of success when challenged in courts. The objects of the delegation reflect distinct invalidation rates, and those rates changed over time. Most notably, early delegation to executive agencies had always fared well in court, and such delegations were even less vulnerable to challenge after 1880 when such cases became extremely common. Delegations to executive branch actors had among the lowest invalidation rates of any type of statutory delegation. Far from being disfavored, as modern proponents of the nondelegation doctrine might suggest, courts seemed to prefer legislative delegations to executive actors during the classical era of nondelegation constitutional jurisprudence.

Table 6: Invalidation Rate in All Cases by Subject Matter of Delegation

Regulation Taxation Spending Other

Total 17% 18% 24% 14% Before 1880 30% 26% 9% 17% After 1880 16% 16% 25% 13%

Examination of delegation by subject matter demonstrates additional implications of nondelegation jurisprudence. Table 6 shows the invalidation rate of statutory provisions in nondelegation cases organized by the subject matter of the delegation. Table 3 shows that a majority of the cases decided 428 University of Pennsylvania Law Review [Vol. 165: 379 after 1880 involved regulatory provisions, in keeping with the growth of cases involving executive agencies. Table 6 indicates that cases involving regulation were also likely to be upheld. Moreover, the invalidation rate of regulatory cases declined dramatically as that type of case became more common after 1880. As Table 6 demonstrates, courts never particularly disfavored delegations of regulatory authority. This is not to say judges never found that legislatures violated the nondelegation principle. The easy case for upholding statutes involved situations where the judge could credibly claim that the legislature had not simply handed over “the power to make a law” but rather had handed down “the power to determine a fact or thing upon which the action of the law depends.”281 But knowing which type of action the legislature had taken might turn simply on how each judge chose to characterize the case. Creating needful rules and regulations was a commonplace job for boards, commissions, and agencies, but the Iowa Supreme Court objected when the legislature authorized the highway commission to make any rules it deemed necessary for traffic safety on state roads. The court thought this was little better than designating the subject matter for commission action, leaving the legislature with nothing to do but “meet and create boards.”282 More often than not, however, judges were content to characterize such laws as complete in themselves, leaving only the details to be fleshed out by administrative agents. Often, the judicial objection to the delegation of power rested on the judicial judgment that the power was not well used. In Ohio, the court feared that the examiner of steam engineers was little more than an “autocrat[] with unlimited discretion,”283 and in Illinois, the court complained that a school board had mandated the vaccination of all school children when “smallpox did not exist in the community, and where there was no cause to apprehend that it was approaching the vicinity of the school, or likely to become prevalent there.”284 And sometimes the legislature simply overstepped the boundaries of what courts could tolerate. The justices of the Maine Supreme Court accepted creation of a commission empowered to make rules and regulations governing fisheries but could not stomach a further provision that the commission’s regulations would preempt any conflicting state statutes.285

281 See State v. Thompson, 60 S.W. 1077, 1079 (Mo. 1901) (en banc) (upholding the power of an auditor to grant gambling licenses to applicants of “good character”). 282 Goodlove v. Logan, 251 N.W. 39, 43 (Iowa 1933). 283 Harmon v. State, 64 N.E. 117, 117 (Ohio 1902) (per curiam). 284 Potts v. Breen, 47 N.E. 81, 85 (Ill. 1897). 285 See McKenney v. Farnsworth, 118 A. 237, 238 (Me. 1922) (holding that “[t]here can be no controversy regarding the unconstitutionality of” a clause that delegates the power to repeal laws). 2017] The Myth of the Nondelegation Doctrine 429

Accordingly, the court ruled that the legislature could not authorize a commission to repeal laws.286 The judges who were deciding nondelegation cases contemporaneously with the growth of the modern administrative state did not seem particularly troubled by those developments. If anything, the courts viewed the delegation of broad regulatory authority to executive agencies with greater favor than earlier innovative delegations to judges or voters. State and federal judges proved willing to accommodate the new institutional devices that emerged in the Gilded Age and expanded in the Progressive Era.

CONCLUSION In a new book on the “unchecked expansion of the state,” former Congressman David McIntosh complains that the courts have gutted the nondelegation doctrine and, in doing so, have abdicated their role of enforcing a strict separation of powers.287 Many conservative critics of the modern administrative state remain haunted by the notion that among the fatalities of the constitutional battles of the New Deal was a robust nondelegation doctrine that imposed significant restraints on the delegation of regulatory authority to agencies and commissions. They claim there was a golden age in which courts stood firm against feckless legislators who sought to pass on hard policy decisions to others, and there were once legal, doctrinal, and political resources for confining the discretion of legislatures to delegate substantive policymaking authority. From this perspective, a serious nondelegation doctrine is part of a constitution-in-exile that can and should be brought in from the cold. This narrative is more mythical than historical. Constitutional lawyers in the nineteenth century understood that the lawmaking power could not be delegated out of the legislative bodies to which the sovereign people had entrusted it. But they also thought that this constitutional commitment posed little obstacle to the rise of the administrative state. The creation of agencies and commissions filled with experts who could effectively make the regulatory policy that shaped the economy was no doubt innovative and required significant rethinking of traditional governmental forms. But state and federal judges did not hesitate to give their stamp of approval to those institutional innovations. Traditional constitutional principles were thought to be capacious enough to accommodate the new administrative structures. That the nondelegation doctrine was more bark than bite in the decades before the New Deal—as it has been in the decades since the New Deal—does

286 Id. 287 David McIntosh & William J. Haun, The Separation of Powers in an Administrative State, in LIBERTY’S NEMESIS: THE UNCHECKED EXPANSION OF THE STATE 239, 239-244 (Dean Reuter & John Yoo eds., 2016). 430 University of Pennsylvania Law Review [Vol. 165: 379 not in itself subvert the normative argument in favor of a more robust nondelegation doctrine. It may well be the case that, as a matter of logic, the modern administrative state is hard to square with foundational constitutional structures. It might be that public policy would be better if crucial decisions now delegated to executive agencies were instead made by legislators. Alternatively, representative democracy might function better if those who stand for election were barred from delegating controversial policy decisions to less accountable government officials. But advocates for such doctrinal reform should be aware that such a confining system of constitutional rules would be an unprecedented change in the American experience, not a return to an earlier governing framework. Judges never developed the sort of doctrinal tools that would allow them to meaningfully distinguish between inappropriate abdication of legislative power and necessary delegation of administrative details. Furthermore, courts never showed the political wherewithal to resist the kinds of administrative innovations that political actors have regarded as necessary for a functional modern government. If there was a falling away from the original constitutional design that allowed for the rise of the modern administrative state, it came well before the New Deal and the constitutional revolution of 1937—even before Woodrow Wilson and the Progressive attack on traditional ideas about the separation of powers.288 In the earliest nineteenth-century nondelegation cases, judges equivocated in the face of the legislative desire to make use of outside parties to flesh out policy directives. Judges have never doubted that there was such a thing as a nondelegation principle that legislatures should recognize and respect, but judges were hesitant to obstruct lawmakers. While considering the work of the United States Sentencing Commission, Justice Scalia lamented, “[W]hile the doctrine of unconstitutional delegation is unquestionably a fundamental element of our constitutional system, it is not an element readily enforceable by the courts.”289 The history of judicial efforts to articulate and apply the nondelegation doctrine would seem to support his assessment.290 Justice Scalia was not the first to worry that the lack of a rigorous nondelegation doctrine could result in the creation of a “junior-varsity Congress”291 that does the lawmaking work of Congress outside the traditional constitutional structure of separated powers. Some have claimed that pre–New Deal judges understood how to prevent that possibility and held the key to preserving the Founders’ original constitutional vision. But,

288 For more discussion on the Progressive theorists of the separation of powers, see JOHN A. ROHR, TO RUN A CONSTITUTION: THE LEGITIMACY OF THE ADMINISTRATIVE STATE 56-89 (1986). 289 Mistretta v. United States, 488 U.S. 361, 415 (1989) (Scalia, J., dissenting). 290 Scalia himself went on to try to identify a bright-line rule that might be enforced by the courts. See id. at 417. 291 Id. at 427. 2017] The Myth of the Nondelegation Doctrine 431 as we have shown, the nondelegation doctrine was not a casualty of the New Deal revolution. Long before then, the nondelegation doctrine—the meaningful and judicially enforced constitutional constraint on legislatures—was already dead.

* * * * *